a bibliometric review of natural language processing

Review ArticleA Bibliometric Review of Natural Language ProcessingEmpowered Mobile Computing

Xieling Chen 1 Ruoyao Ding 2 Kai Xu 3 Shan Wang4

Tianyong Hao 5 and Yi Zhou 6

1College of Economics Jinan University Guangzhou China2School of Information Science and Technology Guangdong University of Foreign Studies Guangzhou China3School of Computer Science and Technology Guangdong University of Technology Guangzhou China4Department of Chinese Language and Literature University of Macau Macau SAR China5School of Computer South China Normal University Guangzhou China6Zhongshan School of Medicine Sun Yat-sen University Guangzhou China

Correspondence should be addressed to Tianyong Hao haotygdufseducn and Yi Zhou zhouyimailsysueducn

Received 23 January 2018 Accepted 5 April 2018 Published 28 June 2018

Academic Editor Javier Prieto

Copyright copy 2018 Xieling Chen et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Natural Language Processing (NLP) empowered mobile computing is the use of NLP techniques in the context of mobileenvironment Research in this field has drawn much attention given the continually increasing number of publications in thelast five years This study presents the status and development trend of the research field through an objective systematicand comprehensive review of relevant publications available from Web of Science Analysis techniques including a descriptivestatistics method a geographic visualization method a social network analysis method a latent dirichlet allocation method andan affinity propagation clustering method are usedWe quantitatively analyze the publications in terms of statistical characteristicsgeographical distribution cooperation relationship and topic discovery and distribution This systematic analysis of the fieldillustrates the publications evolution over time and identifies current research interests and potential directions for future researchOur work can potentially assist researchers in keeping abreast of the research status It can also help monitoring new scientific andtechnological development in the research field

1 Introduction

With the development of mobile devices as well as theadvances in wireless communication technologies mobilecomputing is becoming a significantly important paradigmin todayrsquos world of networked computing systems [1]Mobile computing enables a computer to be used normallywhile in the state of movement Based on perceived situa-tional information in personal and ubiquitous environmentsmobile computing provides services automatically With therapid growth in use of mobile devices far-reaching anddiverse information is being produced rapidly anddistributedinstantly in digitized format [2] A large amount of valu-able information existing in unstructured texts are of greatneed of processing such as web pages short messages

TwitterWeChat messages etc Natural Language Processing(NLP) focuses on the interactions between computers andnatural language texts NLP is capable of providing a com-puter program with the ability to process and understandunstructured texts By automatically analyzing the meaningof user content to take appropriate actions NLP can makeapplications smarter in the mobile environment

NLP empowered mobile computing research field hasattractedmore andmore interests from scientific communitywitnessing 12 publications in 2000 to 55 publications in 2016from Web of Science (WoS) Some representative examplesare as follows Chen et al [3] applied the technique ofmultitask learning using deep neural networks to Mandarin-English code-mixing recognition Three schemes of theauxiliary tasks were proposed to introduce the language

HindawiWireless Communications and Mobile ComputingVolume 2018 Article ID 1827074 21 pageshttpsdoiorg10115520181827074

2 Wireless Communications and Mobile Computing

information to networks and to improve the prediction oflanguage switching for the primary task of senone classifi-cation The proposed schemes enhanced the recognition onboth languages and reduced the relative overall error ratesby 44 on average when dealing with real-world Mandarin-English corpus in mobile voice search Ilayaraja et al [4]presented a weighted association rule mining prefetchingtechnique to determine the secondary service item withthe consideration of access frequency of services semanticdistance among the successive query request and spatialdistance between service instances and user context (egposition service type and query request time) Wong et al[5] analyzed the studentsrsquo vocabulary usage using a corpusanalysis tool to identify and unpack the contextual conditionsin which a mobile- and cloud-assisted Chinese languagelearning environment promoted key learning outcomesRasanen and Saarinen [6] proposed a method based onsparse hyperdimensional coding of sequence structures forsequence prediction Their experiments suggested that themethod was capable of capturing the relevant variable-orderstructure from the sequences ANLP based toolMOTTEwasdeveloped by Puppala et al [7] for extracting and structuringdata in pathology reports automatically to support clinicalsolution applications With an aim of screening informationon human immunodeficiency virusacquired immune defi-ciency syndrome Adesina et al [8] designed a monolingualshort message services based system for the retrieval offrequently asked questions

Bibliometric analysis is defined as the use of statisti-cal methods on evaluating scholarly publications from anobjective and quantitative perspective within a certain field[9] Benefits of bibliometric analysis include (1) organizinginformation in a specific thematic field [10] (2) evaluatingscientific developments in knowledge of a specific subjectand assessing the scientific quality [11] (3) determining theimpact of research funding (4) comparing research perfor-mance across different affiliations and document changes inthe research workforce and (5) identifying emerging areasof research focus and predicting future research success[12] As for researchers especially newcomers bibliometricanalysis can assist them in (1) better selecting potentialresearch topics (2) demonstrating the values and impacts oftheir relevant works (3) recognizing appropriate academicresearchers to seek research collaboration and (4) keepingabreast of new research status and new technological changes[13]

Bibliometric analysis has been widely applied to variousfields for the measurement of quality and productivity ofacademic output and has demonstrated excellent effective-ness from long-term practice Relevant researches mainlyfocused on revealing publication statistical characteristicsexploring the collaboration relationship and uncoveringresearch themes and their evolution Some examples are asfollows Geng et al [14] conducted a bibliometric survey ofthe research field of residential energy and greenhouse gasemissions for the purpose of uncovering research status Intheir work citation analysis was used to assess the influenceof journals countries and authors while network analysiswas performed to evaluate the relationships among countries

authors and keywords Based on 117340 obesity-relatedresearch publications indexed in Scopus database publishedfrom 1993ndash2012 Khan et al [15] reported research trendsand collaboration patterns in the field Roig-Tierno et al[16] conducted a bibliometric analysis on research publica-tions with the application of qualitative comparative analysis(QCA) Their study revealed the differences in quantitativeterms of the three variants of QCA Albort-Morant andRibeiro-Soriano [17] focused on the research developmentof business incubators They sorted 445 publications fromWoS according to bibliographic indicators such as researcharea and year of publication Their study revealed the lackof publications on business incubators and highlighted thefragmented nature of research themes Merigo and Yang[18] aimed at identifying relevant researches and the newesttrends in field of operation research andmanagement scienceThe analysis involved some influential journals two hun-dred most cited publications and productive and influentialauthors Zhang et al [19] quantitatively and qualitativelyevaluated carbon tax related literature from 1989 to 2014using bibliometric analysis Their study demonstrated thatthe USA was the leading country and the Vrije UniversityAmsterdam and Massachusetts Institute of Technology andStanford University were the most productive affiliations inthe research field Randhawa et al [20] conducted a system-atic review of publications on open innovation (OI) researcharea using bibliometrics cocitation analysis and text miningThree distinct areas within OI research were identified iefirm-centric aspects of OI management of OI networks androle of users and communities in OI In order to discover theworldwide trends in the research field of drying bricktileYataganbaba and Kurtbas [21] analyzed relevant patents interms of eg publication number authorship and ownershipand international collaboration patterns Merigo et al [10]explored the research development trends in fuzzy sciencesSimilar works have also been conducted in other fields egnatural language processing [22] neuroimaging [23] anddiabetes [24]

To the best of our knowledge there is no scientificreview of NLP empowered mobile computing research fieldcurrently Thus in this study we conduct a bibliometricanalysis on publications retrieved fromWoS during the years2000ndash2016 to explore the research status of the research fieldThe main objective is to address the following issues (1)investigating publication statistical characteristics and publi-cation collaborations (2) exploring publication geographicaldistributions (3) visualizing scientific collaboration relation-ships and (4) reveling current hot research topic themes andresearch topic changes

The rest of the paper is organized as follows Section 2introduces methods and materials Bibliometric analysisresults on retrieved research publications are reported inSection 3 Findings and discussion are shown in Section 4while Section 5 summarizes the work

2 Methods and Materials

Five different methods are applied to analyze research pub-lications in the NLP empowered mobile computing field

Wireless Communications and Mobile Computing 3

retrieved fromWoSThe details of the methods are describedin Section 21 and the publication data is introduced inSection 22

21 Methods

211 Descriptive Statistics Method Descriptive statistics arebrief descriptive coefficients that summarize a collection ofinformation which can be either a representation of theentire population or a sample Descriptive statistics are com-monly used as measures of central tendency and measuresof variability Measures of central tendency usually includemean median and mode while measures of variability gen-erally contain standard deviation minimum and maximumvariables kurtosis and skewness These two measures usegraphs tables and general discussions to simply describedata This simplifies large amounts of data in a sensible wayby presenting quantitative descriptions in a manageable formto help users understand the meaning of the data beinganalyzed

In this study descriptive statistics method was appliedto acquire characteristics of the retrieved publicationsincluding publication distribution by year most influentialpublications productive journals authors affiliations andcountriesregions as well as co-authors coaffiliation andcocountryregion publication distribution and topic distribu-tion by year

212 Geographic Visualization Method Geographic visual-ization or Geovisualization is a set of tools and techniquessupporting the analysis of geospatial or spatial data empha-sizing knowledge construction over knowledge storage orinformation transmission By combining technologies egimage processing simulation and virtual reality computerscan help present information in a way that patterns canbe found Geovisualization can be applied to all the stagesof problem-solving in geographical analysis from develop-ment of initial hypotheses to knowledge discovery analysispresentation and evaluation According to Toblerrsquos FirstLaw of Geography [25] everything is related to everythingelse but near things are more related than distant thingsThrough Geovisualization we can use location as the keyindex variable and get related informationwhich is previouslyunfound Locations or extents in the earth spacendashtime maybe recorded as datestimes of occurrence Longitude latitudeand elevation are represented as X Y and 119885 coordinatesrespectively

In this study we applied geographic visualization analysisto explore geographical distributions of publications in coun-tryregion level

213 Social Network Analysis Method Social network anal-ysis is a process of investigating social structures usingnetworks and graph theory [26] It focuses on relationshipstructures ranging from casual acquaintance to close bondsNetwork structures are characterized in terms of nodes(items individuals or things within the network) with theedges or links (relationships or interactions) connecting thenodes Researches using social network analysis have been

undertaken in different areas eg collaboration graphs [27]social media networks [28] and disease transmission [29]These networks are often visualized through sociogramsin which nodes are represented as points and edges arerepresented as lines The social network analysis can helpidentify the individuals teams and units who play centralroles leverage peer support and strengthen the efficiency andeffectiveness of existing channels [30]

In this study we applied social network analysis to explorethe cooperation relationships for specific countriesregionsaffiliations and authors in the NLP empowered mobilecomputing research field The cooperation among coun-triesregions affiliations and authors was visualized usinginteractive force directed networks In the networks nodesrepresented specific countriesregions affiliations or authorsand lines indicated cooperation The size of nodes repre-sented publication numbers of a specific country affiliationor author The width of lines reflected cooperation frequen-cies between two countriesregions affiliations or authorsThe color indicated specific continent of a countryregionor specific countryregion of an affiliation or author Userscould explore the cooperation relationships for specific coun-triesregions affiliations or authors by dynamically draggingthe nodes

214 Latent Dirichlet Allocation Method Latent Dirichletallocation (LDA) proposed by Blei [31] is a generativeprobabilistic model The basic idea is that documents arerepresented as random mixtures over latent topics whereeach topic is characterized by a distribution over words andtopics are assumed to be uncorrelated

LDA formally defines the following terms

(1) A word is defined as an item from a vocabularyindexed by 1 119881

(2) A document is a sequence of119873 words denoted by 119889 =(1199081 119908119873)(3) A corpus is a collection of119872 documents denoted by119863 = 1198891 119889119872LDA assumes the following generation process

(1) The term distribution 120573 which contains the probabil-ity of a word occurring in a given topic is determinedby 120573 sim Dirichlet(120575)

(2) The proportions 120579 of the topic distribution for adocument 119889 are determined by 120579 sim Dirichlet(120572)

(3) For each word119908119894 in the document d a topic is chosenby the distribution 119911119894 sim Multinomial(120579) and a wordis chosen from amultinomial probability distributionconditioned on the topic 119911119894 119901(119908119894 | 119911119894 120573)

As for variational expectation-maximization (VEM) esti-mation the log-likelihood for one document 119889 isin 119863 is givenby

ℓ (120572 120573) = log (119901 (119889 | 120572 120573))


= logintsum119911

[ 119873prod119894=1

119901 (119908119894 | 119911119894 120573) 119901 (119911119894 | 120579)]sdot 119901 (120579 | 120572) 119889120579

(1)

Gibbs sampling defines a Markov chain in the spaceof possible variable assignments such that the stationarydistribution of theMarkov chain is the joint distribution overvariables Thus it is a Markov Chain Monte Carlo method[32] Its aim is to construct a Markov chain converging tothe target probability distribution in the high dimensionalmodel and then the sample distribution closest to the targetprobability distribution will be extracted The log-likelihoodfor Gibbs sampling can be obtained through

log (119901 (119889 | 119911)) = 119896 log(Γ (119881120575)Γ (120575)119881)

+ 119896sum119870=1

[[119881sum119895=1

log (Γ (119899(119895)119870 + 120575))]]minus log (Γ (119899(sdot)119870 + 119881120575))

(2)

The perplexity as shown in (3) is often used to evaluatethe models on held-out data and is equivalent to the geomet-ric mean per-word likelihood The less the perplexity is thebetter the model is

perplexity (119889) = expminus log (119901 (119889))sum119863119889=1sum119881119895=1 119899(119895119889)

(3)

In (4) 119899(119895119889) denotes how often the jth term occurs in thedth document If the model is fitted through Gibbs samplingthe likelihood can be determined for the perplexity using

log (119901 (119889)) = 119863sum119889=1

119881sum119895=1

119899(119895119889) log[ 119896sum119870=1

120579(119889)119870 120573(119895)119870 ] (4)

Additionally estimation using Gibbs sampling requiresspecification of values for the parameters of the prior distri-butions

In this study topic discovery and distribution wereanalyzed using LDA models with the following steps

(1) We assigned the weights of segmented author key-words and Keywords Plus publication title andabstract as 04 04 and 02 respectively as determinedin our former experiment [13]

(2) Term Frequency-Inverse Document Frequencies(TF-IDF) were used to filter out unimportant termsAs one of the most popular term-weighting schemesTF-IDF increases proportionally to the numberof times a term appears in a publication but isoften offset by the frequency of the term in the

whole collection of publications We calculated theTF-IDF values of all terms to sort the terms Bymanually examining these ranked terms we defineda threshold as 01 empirically Only the terms with aTF-IDF value greater than the threshold were keptfor further analysis

(3) Through sampling 16 different topic numbers wereset to 119888(2 10 15 20 40 50 80 150 250) For eachtopic number 10-fold cross-validation was used toevaluate model performance Specifically dataset wassplit into 10 test datasets to conduct multiple runsPerplexity criteria were used to select optimal topicnumber 120572 for Gibbs sampling was initialized as themean value of 120572 values for model fitting using VEMwith the optimal topic number

(4) With an initialized 120572 and the optimal topic numberwe adopted Gibbs sampling and VEM method toestimate the LDA model

(5) By matching the topics detected by VEM andGibbs sampling based on Hellinger distance thebest matches with the smallest distance could beidentified Hellinger distance is calculated as (5) inwhich 119875 and 119876 denote two probability measures

1198672 (119875 119876) = 12 int (radic119889119875 minus radic119889119876)2 (5)

215 Affinity Propagation ClusteringMethod Affinity Propa-gation (AP) algorithm was proposed by Frey and Dueck [33]It is a technique for data clustering based onmessage passingAP does not require the predefined number of clustersIt identifies cluster centers or exemplars as representativemembers of clusters Initially all nodes are considered asexemplars ldquoPreferencerdquo is used to reflect how likely one nodeis chosen as an exemplar If no prior knowledge is availableall nodes will be assigned the same preference value APhas been shown to be more efficient and effective in clusteridentification than traditional clustering methods eg 119896-means [34]

AP algorithm takes 119904(119894 119895) as function of similarity toreflect the fitness of the data point 119895 being the exemplar ofdata point 119894 The aim of AP is to maximize the similarity119904(119894 119895) between every data point 119894 and its chosen exemplar119895 Each node 119894 also has a self-similarity 119904(119894 119894) Individualdata points initialized with a larger self-similarity are morelikely to become exemplars All data points are equally likelyto be exemplars when they are initialized with the sameconstant self-similarityThe number of clusters produced willbe increased and decreased accordingly with this commonself-similarity input

There are two types of messages contained in this tech-niqueThe responsibility 119903(119894 119895) is directed from 119894 to candidateexemplar 119895 It indicates how well suited 119895 is to be 119894rsquos exemplartaking into consideration competing potential exemplarsThe availability 119886(119894 119895) is sent from candidate exemplar 119895 backto 119894 It indicates 119895rsquos desire to be an exemplar for 119894 basedon supporting feedback from other data points Both theself-responsibility 119903(119894 119894) and the self-availability 119886(119894 119894) can


Table 1 The query used to retrieve research publications in the NLP empowered mobile computing field fromWoS

TS=((ldquonatural language processingrdquo OR ldquoNLPrdquo OR ldquosemantic analysisrdquo OR ldquobag of wordsrdquo OR ldquoword sense disambiguationrdquo OR ldquonamedentity recognitionrdquo OR ldquoNERrdquo OR ldquosentiment analysisrdquo OR ldquoinformation extractionrdquo ldquotokenizationrdquo OR ldquostemmingrdquo ORldquolemmatizationrdquo OR ldquocorpusrdquo OR ldquostop wordsrdquo OR ldquoparts-of-speechrdquo OR ldquolanguage modelingrdquo OR ldquon-gramsrdquo OR ldquosyntactic analysisrdquoOR ldquoinformation retrievalrdquo OR ldquolanguage modelrdquo) AND (ldquomobile computingrdquo OR ldquomobilerdquo OR ldquosmart devicerdquo OR ldquosmartphonerdquo ORldquocellphonerdquo OR ldquotelephony devicerdquo OR ldquoCellular networkrdquo OR ldquoAndroidrdquo OR ldquoiOSrdquo OR ldquophonerdquo))

reflect accumulated evidence that 119894 is an exemplarThe updateformulas for responsibility and availability are as follows

119903 (119894 119895) larr997888 119904 (119894 119895) minus max1198951015840st 1198951015840 =119895

119886 (119894 1198951015840) + 119904 (119894 1198951015840)

119886 (119894 119895)119894 =119895

larr997888 min0 119903 (119894 119895) + sum

forall1198941015840notin119894119895

max 0 119903(1198941015840 119895119886 (119895 119895) larr997888 sum

1198941015840st 1198941015840 =119895max 0 119903 (1198941015840 119895)

(6)

Responsibility and availability of message updates are119898new = 120582119898old + (1 minus 120582)119898new where 120582 is a weightingfactor between 0 and 1 In AP the clustering is completewhen the messages converge Also AP algorithm is ableto determine when a specific data point has converged tocluster head status in its given cluster A point becomes thecluster head when its self-responsibility plus self-availabilitybecomes positive Upon convergence each node 119894rsquos clusterhead can be calculated using

119862119867119894 = argmax119895119886 (119894 119895) + 119903 (119894 119895) (7)

In our study with the basis of term-topic posteriorprobability matrix we applied AP clustering method for thecluster analysis of the topics identified by the LDA method

22 Materials Web of Science as the most authoritativecitation database was used as the data source for retrievingresearch publications in the NLP empowered mobile com-puting field First of all a list of keywords related to theldquonatural language processingrdquo and ldquomobile computingrdquo wasdetermined by a domain expertWith ldquoScienceCitation IndexExpandedrdquo and ldquoSocial Sciences Citation Indexrdquo as indexespublications used in this study were identified using thespecific query in Table 1 716 publications in ldquoarticlerdquo typeduring years 2000ndash2016 were obtained Citations counted toSeptember 8th 2017 were considered for each publication

The raw data of the 716 publications were downloadedas plain text Key elements including title author journalpublication date subject category language funding authorkeywords Keywords Plus abstract and author address aswell as number of citations pages and references wereextracted In order to ensure they were closely related tothe research field manual verification was conducted by adomain expert on each publication 471 publications wereidentified as relevant for analysis eventually Further cor-responding affiliations and countriesregions were identi-fied out from author address information Key terms were

extracted from author keywords Keywords Plus title andabstract

The statistical characteristics of the publications areshown as Table 2 The average page number of the pub-lications is 1566 and the average reference number of thepublications is 3329There are 48 subject categories includedwhere the top 3 categories are computer science (3876)engineering (1627) and telecommunications (1098)

The distribution characteristics of the 471 publicationsare shown in Figure 1 Figure 1(a) shows the distributions ofthe numbers of countriesregions affiliations authors andfunds Figure 1(b) shows the distributions of the numbersof keywords pages and references The distribution ofthe number of title characters is shown in Figure 1(c) InFigure 1(d) the right bottom illustrates the distribution of thenumber of abstract characters

3 Results

31 Publication with Year The total publications total cita-tions average number of citations per publication and thenumber of annual citations are demonstrated in Figure 2The results show that the research in the NLP empoweredmobile computing field exhibits an overall upward trend influctuation (from 12 publications in 2000 to 55 publicationsin 2016)The publication number presents a stable increasingtrend since 2010 Based on the data for years 2010ndash2016 wedeveloped a regression model by setting the independentvariables as time1000 and (time1000)2The estimated regres-sion model is calculated as 119910 = 67143 lowast 103 minus 134777 lowast104119909 The adjusted goodness-of-fit 1198772 of the model is 09468With the regression model publication number in 2017 ispredicted as 65 while the actual number of publications onWoS in 2017 is 66 The trend of citations does not keepstep with publication number and extreme values appearin 2002 as 431 2007 as 503 and 2010 as 490 The averagenumber of citations per publication is calculated as totalcitationstotal publications It shows an overall downwardtrend in fluctuation from 2192 in 2000 to 253 in 2016 Weeliminated the influence of duration since first publicationusing the formula the number of annual citations (CY)= total citations(2016 + 1-publishing year) The number ofannual citations increases in fluctuation from 1547 in 2000to 139 in 2016

32 Productive Journals The top 11 contributing journals inthe research field are presented in Table 3 These journalscontribute about 21 of the total publications and 2920of the total citations The most productive 3 are IEEEACMTransactions on Audio Speech and Language Processing (25


Table 2 The statistical characteristics of the 471 publications

Characteristics StatisticsTotal pub 471pub with author keywords or Keywords Plus 412unique publication sources 287unique countries (or regions)first countries (or regions) 60 52unique affiliationsfirst affiliations 544 345unique authorsfirst authorslast authors 1408 451 441Average citations 1042Average countries (or regions) in one pub 125Average affiliations in one pub 164Average authors in one pub 327Average funds in one pub 073Average pages in one pub 1566Average references in one pub 3329Average author keywords or Keywords Plus 681Average wordscharacters in title 1057 8013Average wordscharacters in abstract 18640 126558

Language distribution English (9873) Estonian (042) French (042) Spanish (021)Afrikaans (021)

Subject category distribution (Top 10)Computer Science (3876) Engineering (1627) Telecommunications

(1098) Acoustics (582) Information Science amp Library Science (278)Linguistics (251) Psychology (212) Operations Research amp ManagementScience (185) Business amp Economics (132) Communication (132)

Top 10 terms in author keywords and Keywords PlusMobile (3036) Information (2208) Retrieval (1677) Recognition(1656) System (1486) Speech (1401) Model (1210) Network

(1210) Language (1104) Analysis (998)

Top 10 terms in titlesMobile (3418) Information (1783) System (1253) Retrieval (1210)Recognition (1062) Speech (870) Network (828) Model (786)

Language (722) Environment (637)

Top 10 terms in abstractsMobile (6667) Information (5690) Paper (5541) System (4820)Result (4607) Data (3800) User (3800) Model (3737) Device

(3270) Retrieval (3142)

Table 3 Top 11 contributing journals in the NLP empowered mobile computing research field

Rank Journals SC TP P TC ACP 119867 ge10 T1001 IEEEACM Transactions on Audio Speech and Language Processing A E 25 531 447 1788 11 12 112 Speech Communication A CS 11 234 179 1627 6 6 53 Computer Speech and Language CS 10 212 93 930 6 5 34 Expert Systems with Applications CS E ORampMS 8 170 320 4000 8 7 54 IEEE Transactions on Consumer Electronics E T 8 170 44 550 5 1 06 Mobile Information Systems CS T 7 149 95 1357 3 2 26 Multimedia Tools and Applications CS E 7 149 71 1014 3 1 16 Personal and Ubiquitous Computing CS T 7 149 67 957 4 3 19 Information Sciences CS 6 127 85 1417 5 3 310 EURASIP Journal on Wireless Communications and Networking E T 5 106 22 440 2 1 110 IEICE Transactions on Information and Systems CS 5 106 11 220 2 0 0Notice Journal IEEETransactions onAudio Speech and Language Processing changed name as IEEEACMTransactions onAudio Speech and Language Processingin 2013 and journal IEEETransactions on Speech andAudio Processing ceased publication in 2005 and the current retitled publication is IEEEACMTransactionson Audio Speech and Language Processing Therefore publications from these 2 journals were combined as published by IEEEACM Transactions on AudioSpeech and Language ProcessingAbbreviations SC subject categories only with NLP empoweredmobile computing research (A acoustics E engineering CScomputer science ORampMS operations research andmanagement science T telecommunications) TP total publications 119875 percentage of the publicationsTC total citations ACP average number of citations per publication calculated as TCTP H H-index ge10 number of publications with citations ge10 T100number of publications in the top 100 most influential publications


(a) (b)

(c) (d)

Figure 1 Distribution characteristics of the 471 publications

055110165220275330385440495550

0153045607590

105120135150

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

cita

tion

pub

av

g

cita

tion

per p

ub

and

ann

ual c

itatio

n

pubcitation

avg citationannual citation

Figure 2The statistics of the 417 publications (the light blue bars indicate total publications and the red bars indicate total citationsThe darkblue line indicates average citations per publication and the green line indicates annual citations)


Speech Commun

Comput Speech Lang

IEEE-ACM Trans AudioSpeech Lang

Expert Syst Appl

IEEE

Electron

Trans Consum

Mob Inf Syst

Multimed Tools Appl

Pers UbiquitousComput

Inf Sci EURASIP J Wirel

Commun

IF 1768 19 3928 1694 0849 153 2395 4832 1529 0411SJR 0593 0536 1433 0587 0219 0396 0648 191 0304 01915-Year IF 1769 1938 3526 1725 1 1572 2512 4731 1558 0332SNIP 1628 175 2492 1447 0509 0775 1651 2537 0698 0419CiteScore

2491081325013143

35 238 267 47 256 103 119 312 537 136 049

IEICE Trans Inf

Syst

0

05

1

15

2

25

3

35

4

45

5

Figure 3 Comparisons of IF SJR 5-Year IF SNIP and CiteScore for the top 11 productive journals for year 2016

publications 447 citations 1788ACP and 11119867-index) SpeechCommunication (11 publications 179 citations 1627ACP 6119867-index) and Computer Speech and Language (10 publica-tions 93 citations 930 ACP 6119867-index) Expert Systems withApplications has the highest ACP of 4000 We found that32 of the 100 most influential publications are published inthe 11 journals According to subject category of these 11journals computer science possesses the widest influence inthe research field

In order to better measure the overall scientific impor-tance of these 11 journals 5 assessment indicators acquiredfromScientific Journal Rankingswere used including ImpactFactor (IF) SCImago Journal Rank (SJR) 5-Year IF SourceNormalized Impact per Paper (SNIP) and CiteScore IFis a measure for reflecting the yearly average number ofcitations to recent publications published in a journal Itis the primary and widely used indicator on assessing onejournalrsquos significance SJR is a measure of scientific influenceof scholarly journals It accounts for both the number ofcitations received by a journal and the importance or prestigeof the journals where such citations come from 5-Year IF iscalculated by dividing the number of citations to the journalin a given year by the number of publications published inthat journal in the previous five years SNIP is defined asthe ratio of the journalrsquos citation count per publication andthe citation potential in its subject field CiteScore indexlaunched by Elsevier in December 2016 is calculated asthe ratio of total citations received in a given year by allpublications published in a given journal in three previous

years and the number of publications published in the journalin three previous years

Therefore the 11 productive journals were compared byusing their IF SJR 5-Year IF SNIP and CiteScore for year2016 as shown in Figure 3 As for IF SJR and CiteScorethe top 3 are Information Sciences (IF 4832 SJR 191 andCiteScore 537) Expert Systems with Applications (IF 3928SJR 1433 and CiteScore 47) and IEEEACMTransactions onAudio Speech and Language Processing (IF 2491 SJR 0813and CiteScore 35) As for 5-Year IF the top 3 are InformationSciences (5-Year IF 4731)Expert SystemswithApplications (5-Year IF 3526) and Personal and Ubiquitous Computing (5-Year IF 2512) As for SNIP score the top 3 are IEEEACMTransactions on Audio Speech and Language Processing (SNIP3143) Information Sciences (SNIP 2537) and Expert Systemswith Applications (SNIP 2492)

33 Most Influential Publications The number of citationsreflects the popularity and influence of a publication in thescientific community [10]Thus we used the total citations asameasurement of influenceThere are 69 and 129 publicationswith the number of citations ge20 and ge10 Top 15 mostinfluential publications are listed in Table 4 The publicationby Miao et al [35] in 2010 (376 citations) is the mostinfluential one followed by [36] published by MacKenzieand Soukoreff in 2002 (172 citations) and [37] by Strayerand Drews in 2007 (148 citations) We further consider thenumber of annual citations of the 15 publications The top 3publications measured by this indicator are [38] by Cao et al


Table 4 Top 15 most influential publications in the NLP empowered mobile computing research field

Rank Title Authors Year TC CY1 Energy-Efficient Link Adaptation in Frequency-Selective Channels Miao G W et al 2010 376 5371

2 Text Entry for Mobile Computing Models and Methods Theory andPractice MacKenzie I S Soukoreff R W 2002 172 1147

3 Cell-Phone-Induced Driver Distraction Strayer D L Drews F A 2007 148 1480

4 A Vector Space Modeling Approach to Spoken LanguageIdentification Li H Z et al 2007 116 1160

5 Context-Aware System for Proactive Personalized Service Based onContext History Hong J Y et al 2009 91 1138

6 More than Words Social Networksrsquo Text Mining for Consumer BrandSentiments Mostafa M M 2013 88 2200

7The Effect of Mobility-Induced Location Errors on GeographicRouting in Mobile Ad Hoc and Sensor Networks Analysis andImprovement Using Mobility Prediction

Son D J et al 2004 77 592

8 A Personalized Tourist Trip Design Algorithm for Mobile TouristGuides Souffriau W et al 2008 76 844

9 DrsquoAgents Applications and Performance of a Mobile-Agent System Gray R S et al 2002 73 487

10 Optical Encryption and QR Codes Secure and Noise-FreeInformation Retrieval Barrera J F et al 2013 64 1600

11 Text-Dependent Speaker Verification Classifiers Databases andRSR2015 Larcher A et al 2014 60 2000

12 A Location-Aware Recommender System for Mobile ShoppingEnvironments Yang W S et al 2008 59 656

12 An Application of Reinforcement Learning to Dialogue StrategySelection in a Spoken Dialogue System for Email Walker M A 2000 59 347

14 Landmark Recognition with Compact BoW Histogram andEnsemble ELM Cao J W et al 2016 56 5600

14 Mobile-Agent Coordination Models for Internet Applications Cabri G et al 2000 56 329Abbreviations TC total number of citations during 2000 and 2016 CY the number of annual citations

published in 2015 (119862119884 = 56) [35] by Miao et al in 2010(119862119884 = 5371) and [39] byMostafa in 2013 (119862119884 = 22)These3 publications rank 14th 1st and 6th respectively accordingto total citations

34 Productive Authors and Affiliations From the 471 pub-lications there are 1408 authors 451 of them are firstauthors and 441 are last authors 20 authors have 3 or morepublications and 98 authors have 2 or more publications20 most productive authors are listed in Table 5 Accordingto the result the most productive authors are Chen Taofrom Singapore (4 publications supported by 4 funds 108citations 27 ACP and 4 119867-index) and Mizzaro Stefanofrom Italy (4 publications 45 citations 1125 ACP and 3 119867-index) Chen Tao is listed as first author of 3 publicationsand all the 3 publications appear in top 100 most influentialpublications Mizzaro Stefano cooperates with others in allhis 4 publications and 1 publication appears in the top 100As for the ranking based on citation number the top 3productive authors are Lee Chin-Hui from the USA (173citations and 5767 ACP) Chen Tao from Singapore (108citations and 27 ACP) and Xie Xing from China (51 citationsand 17 ACP) Ranking based on the ACP indicator yields thesame result Kim Harksoo from South Korea achieves themost funding supports ie 7 for his 3 publications

544 affiliations from 60 countriesregions have publica-tions in the NLP empowered mobile computing researchfield Table 6 lists 15 most productive affiliations Amongthem 5 are from the USA 3 from China 2 from Taiwan 1from India 1 from Italy 1 fromSouthKorea 1 fromSingaporeand 1 from England The top 4 most productive affiliationsa Nanyang Technological University from Singapore (8 pub-lications 87 citations 1088 ACP and 5 119867-index) TsinghuaUniversity from China (8 publications 42 citations 525 ACPand 4 119867-index) Microsoft Research Asia from China (7publications 115 citations 1643 ACP and 5 119867-index) andNational Taiwan University from Taiwan (7 publications 83citations 1186 ACP and 5 119867-index) Nanyang TechnologicalUniversity cooperates with others in 5 publications and servesas first affiliation in 4 of them 3 of these 5 publications appearin the list of top 100 most influential publications TsinghuaUniversity cooperates with others in 4 publications and servesas first affiliation in all 8 publications These 8 publicationsare supported by 21 funds As for the ranking based on thetotal citations the top 3 are Georgia Institute of Technologyfrom theUSA (550 citations and 110 ACP)Microsoft ResearchAsia from China (115 citations and 1643 ACP) and NationalCheng Kung University from Taiwan (62 citations and 124ACP) Ranking based on the ACP indicator yields the sameresult


Table 5 The most productive authors in the NLP empowered mobile computing research field

Rank Name Country TP TC ACP 119867 T100 119865 FP LP CP1 Chen Tao SG 4 108 2700 4 3 4 3 0 41 Mizzaro Stefano IT 4 45 1125 3 1 0 0 0 42 Baek Jin-Wook KR 3 27 900 3 1 0 1 0 32 Bertino Elisa USA 3 40 1333 3 1 3 0 2 32 Cacciapuoti Angela Sara IT 3 18 600 2 1 4 3 0 32 Caleffi Marcello IT 3 18 600 2 1 4 0 2 32 Christodoulakis Stavros GR 3 9 300 1 0 0 0 3 32 Crestani F UK 3 37 1233 2 1 0 0 3 32 Jung Jason J KR 3 4 133 1 0 2 0 2 32 Karanastasi Anastasia GR 3 9 300 1 0 0 1 0 32 Kazasis Fotis G GR 3 9 300 1 0 0 0 0 32 Kim Harksoo KR 3 4 133 1 0 7 0 2 32 Lee Chin-Hui USA 3 173 5767 3 3 4 1 2 22 Liu Jia CN 3 3 100 1 0 6 0 1 32 Muneyasu Mitsuji JP 3 4 133 2 0 2 1 2 32 Pierre Samuel CA 3 38 1267 2 1 0 0 0 32 Seide Frank CN 3 48 1600 2 2 0 0 0 22 Xie Xing CN 3 51 1700 3 2 3 1 0 32 Yan Yonghong CN 3 9 300 2 0 3 0 3 32 Yeom Heon Y KR 3 27 900 3 1 0 0 3 3Abbreviations CACanadaUSA theUSAUK England CNChina KR SouthKoreaGRGreece IT Italy JP Japan SG Singapore TP total publications TCtotal citations ACP average number of citations per publication119867119867-index T100 number of publications in the top 100 highly cited publications 119865 numberof publications with funding FP number of publications as first author LP number of publications as last author CP number of collaborated publications

Table 6 The most productive affiliations in the NLP empowered mobile computing research field

Rank Name Country TP TC ACP 119867 T100 119865 FP CP1 Nanyang Technological University SG 8 87 1088 5 3 1 4 51 Tsinghua University CN 8 42 525 4 1 21 8 43 Microsoft Research Asia CN 7 115 1643 5 4 3 3 63 National Taiwan University TW 7 83 1186 5 3 3 5 45 Georgia Institute of Technology USA 5 550 11000 4 4 6 2 45 Massachusetts Institute of Technology USA 5 10 200 2 0 9 4 45 National Cheng Kung University TW 5 62 1240 3 2 6 4 15 Purdue University USA 5 47 940 4 1 4 1 59 Indian Institute of Technology IN 4 35 875 3 1 3 4 19 Microsoft Corporation USA 4 28 700 3 1 1 4 29 The Pennsylvania State University USA 4 26 650 3 0 8 2 29 Seoul National University KR 4 31 775 4 1 5 4 09 University of Strathclyde UK 4 43 1075 2 1 0 4 09 University of Udine IT 4 45 1125 3 1 0 1 39 Zhejiang University CN 4 43 1075 2 1 5 3 3Abbreviations USA the USA UK England CN China SG Singapore TW Taiwan IN India KR South Korea IT Italy TP total publications TC totalcitations ACP average number of citations per publication119867119867-index T100 number of publications in the top 100 highly cited publications 119865 number ofpublications with funding FP number of publications as first affiliation CP number of collaborated publications

35 Geographical Distribution The 471 publications are from60 countriesregions The number of publications affiliatedwith 1 countryregion range [61 105) 3 countriesregionsrange [37 61) and 5 range [11 17) Table 7 shows top 15 mostproductive countriesregions in the field Figure 4 illustratesgeographical distributions of the publications The top 4

countries are the USA (105 publications 1795 citations 171ACP and 22119867-index) China (61 publications 372 citations61 ACP and 10 119867-index) England (44 publications 418citations 95 ACP and 12 119867-index) and South Korea (41publications 281 citations 685 ACP and 8119867-index) Amongthe 105 publications from the USA 32 appear in the list of top


Table 7 The most productive countriesregions in the NLP empowered mobile computing research field

Rank Country TP TC ACP 119867 T100 FP () Single-countryregion International collaborationACP TP () ACP TFC (119899)

1 USA 105 1795 1710 22 32 7714 2078 6000 1157 CN (12)2 CN 61 372 610 10 10 9180 417 5738 904 USA (12)3 UK 44 418 950 12 11 6136 1168 6364 569 IECH (2)4 KR 41 281 685 8 6 9268 703 8537 583 CNUSA (3)5 TW 37 399 1078 11 11 9459 1107 8108 957 USA (4)6 JP 24 77 321 3 1 7917 144 7500 850 CN (3)7 IT 21 299 1424 10 9 8095 1319 7619 1760 USA (3)8 AU 18 218 1211 7 7 6111 1800 3889 836 USA (5)8 CA 18 313 1739 9 4 8889 2038 7222 960 NA10 FR 17 157 924 6 5 6471 445 6471 1800 CNUSA (2)10 GR 17 38 224 3 0 10000 224 10000 000 NA10 ES 17 124 729 7 2 8824 643 8235 1133 USA (2)13 SG 16 355 2219 9 7 7500 1490 6250 3433 USA (2)14 HK SAR 15 98 653 6 2 5333 917 4000 478 CNUSA (4)15 DE 14 114 814 5 3 8571 811 6429 820 CN (12)Abbreviations USA America UK England CN China SG Singapore TW Taiwan KR South Korea IT Italy JP Japan AU Australia CA Canada FRFrance GR Greece ES Spain HK SAR Hong Kong SAR DE Germany IE Ireland CH Switzerland TP total publications TC total citations ACP averagenumber of citations per publication119867119867-index T100 number of publications in the top 100 highly cited publications FP () percentage of publications asfirst affiliation TFC (119899) number of cooperation times with the closest collaborator where 119899ge 2

lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000

Figure 4Geographical distributions of theNLP empoweredmobilecomputing research publications

100 most influential publications It is noted that publicationsfrom Singapore have the highest ACP which indicates thehigh quality of the publications As for most of the top 15productive countriesregions the international collaborationrates are around 30 except for Greece with 0 and Australiawith 6111 The USA is the closest collaborator for 9 of the15 countriesregionsTheACP of internationally collaboratedpublications is much higher than that of noninternationallycollaborated publications for countriesregions like ChinaJapan Italy France Spain and Singapore This potentiallyindicates that international collaboration can improve thequality of their publications

Since the publications are mainly distributed in the USAChina England and South Korea we further explored theannual publication distributions for these 4 countries asshown in Figure 5 The number of publications for the USAand China is on the whole presenting upward trend influctuation As for the USA the number increases from 2in 2000 to 9 in 2007 but dwindles to 2 in 2010 After thatthe upward trend becomes more significantThe situation forChina is quite like that for the USA after 2010 witnessingthe great mass upsurge on the NLP empowered mobilecomputing research in these two countries since 2010 As forEngland and South Korea the number of publications doesnot increase much in fluctuation with years going on

36 Cooperation Relationship Figure 6 shows the trendsof the international collaborative and the percentage ofinternational collaborative publications We found that theinternational collaborative publications increase during theyears 2000ndash2016 The percentage of international collabo-rations increases from 833 in 2000 to 3273 in 2016This indicates that international collaborations in the NLPempowered mobile computing research field have becomeincreasingly important

Figures 7 and 8 present the institutional level of coop-eration and the author level of cooperation respectivelyThe cooperation between different institutions is becomingmore and more frequent The percentage of institution-collaborative publication increases from 1667 in 2000 to5818 in 2016 More than 90 of the publications are mul-tiauthored since 2011 It is worth noticing that the percentagereaches up to 100 in 2015

Furthermore the cooperation relations for specific coun-triesregions affiliations and authors were visualized with


Figure 5 Publication distributions by year for the top 4 countriesregions

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

international-collaborative pubPercentage of international-collaborative pub

0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)

Figure 6 International collaborative publication distribution by year

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

institution-collaborative pubPercentage of institution-collaborative pub

0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub

Figure 7 Institution-collaborative publication distribution by year


Figure 8 Author-collaborative publication distribution by year

USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria

Bangladesh Bosnia amp Herceg

Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand

Figure 9 Cooperation network of 48 countriesregions (node colors represent different continents eg orange for Asia blue for NorthAmerica green for Europe red for Oceania purple for Africa and brown for South America) The network can be accessed via the link(httpwwwzhukunorghaotyresourcesaspid=NLPEMC cocountry)

social network analysis A cooperation network for 48 coun-triesregions is shown in Figure 9 17 of them come fromAsia (represented as orange nodes) 3 from North America(represented as blue nodes) 22 from Europe (represented asgreen nodes) 3 from Africa (represented as purple nodes)2 from South America (represented as brown nodes) and 1from Oceania (represented as red node) There are 141 affili-ations with the number of publications ge 2 and there existscooperation among 91 of them Figure 10 shows a cooperation

network of the 91 affiliations 23 of the 91 affiliations arefrom the USA and 14 from China As for cooperation ofauthor level there are 98 authors with publication count ge 2among them 65 authors involve in cooperationWe created acooperation network of the 65 authors as shown in Figure 11

37 Topic Discovery and Distribution By setting TF-IDFvalue threshold as 01 the terms were ranked by frequencyTable 8 lists top 20 most frequent terms in which the top 5


Aalto Univ

ASTAR

Beihang Univ

Carnegie Mellon Univ

Chaoyang Univ Technol

China Univ Geosci

Chinese Univ Hong Kong

Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ

Columbia UnivDalian Maritime Univ

Dalian Univ Technol

Fdn Ugo Bordoni

Georgia Inst Technol

Gyeonggi Collage Sci amp Technol

Haukeland Hosp

Hong Kong Univ Sci amp Technol

Inha Univ

Inst Mines Telecom SudParis

Int Inst Informat Technol

Intel Corp

Kangwon Natl Univ

King Abdulaziz City Sci amp Technol

King Saud Univ

Korea Adv Inst Sci amp TechnolKorea Univ

Kwangwoon Univ

Kyungpook Natl U

Microsoft CorpMicrosoft Res Asia

Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ

Natl Chiao Tung Univ

Natl Pingtung Univ Sci amp T

Natl Sun Yat Sen Univ

Natl Taipei Univ

Natl Taiwan Univ Sci amp Technol

Natl Taiwan Univ Natl Tsing Hua Univ

Natl Univ Def Technol

Natl Univ Singapore

North West Univ

Northwestern Polytech Univ

Norwegian Univ Sci amp Technol

Oakland Univ

Ohio State UnivPeking Univ

Penn State Univ

Purdue Univ

Queensland Univ Technol

Renmin Univ China

Sogang Univ

SRI Int

Stanford UnivTexas Tech Univ

Tsinghua UnivTunghai Univ

UCL

Univ Athens

Univ

Univ Calif Berkeley

Univ Carlos III Madrid

Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina

Univ Sci amp Technol China

Univ Sheffield

Univ SUniv Texas Dallas

Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington

Univ Wisconsin Madison

Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ

Figure 10 Cooperation network of 91 affiliations (node colors represent different countriesregions eg red for the USA pink for SouthKorea and purple forAustralia)Thenetwork can be accessed via the link (httpwwwzhukunorghaotyresourcesaspid=NLPEMC coaffil-iation)

terms are ldquoAgentrdquo (369) ldquoImagerdquo (215) ldquoSentimentrdquo (128)ldquoDialoguerdquo (83) and ldquoHealthrdquo (81) Figure 12 presents theperplexities of models fitted by using Gibbs sampling withdifferent numbers of topics The result suggests that theoptimal topic number is between 40 and 80 Hence weset the topic number as 40 The 120572 was set to the meanvalue 001101332 in the cross-validation fitted using VEMUsing the parameters we estimated the LDA model usingGibbs sampling By semantics analysis of representative termsin each topic as well as reviewing text intention of thecorresponding publications we assigned potential theme toeach topic The order of topics are determined based onHellinger distance Specifically Topic 36 is the best matchingtopic and Topic 11 ranks 2nd while Topic 37 is the worsematching one Due to space limitation Table 9 only displaysthe top 10 best matching topics with the most frequentterms Each publication was assigned to the most likelytopic with the highest posterior probability Integrating topicproportions for all the publications we obtained a topicdistribution The 4 most frequent research topics are Topic36 (638) Topic 4 (426) Topic 11 (383) and Topic 17(383) while the 4 least frequent research topics are Topic

26 (149) Topic 23 (128) Topic 10 (106) and Topic 20(106)

We used the AP clustering analysis to perform the clusteranalysis of the 40 topics One way for measuring topic sim-ilarity is based on term-level similarity with the hypothesisthat topics may contain the same terms The clustering resultbased on term-topic posterior probability matrix is shown inFigure 13 where the 40 topics are categorized into 8 groups

Identifying emerging research topics can provide valuableinsights into the development of the research field Like-wise identification of fading research topics can also helpunderstand the hot spots evolution [40] We then exploredthe annual publication proportions of the 40 research topicsas shown in Figure 14 We used MannndashKendall test [41] anonparametric trend test to examine whether increasing ordecreasing trends are existing in the 40 topics Test resultsshow that 12 topics including Topic 1 Topic 4 Topic 7 Topic10 Topic 14 Topic 18 Topic 20 Topic 26 Topic 29 Topic32 Topic 33 and Topic 39 present a statistically significantincreasing trend While Topic 36 presents a statisticallysignificant decreasing trend both at the two-sided 119901 = 005levels


Alkhairy Ashraf

Baek Jin-Wook

Barnard EtienneBertino Elisa

Bless Josef J

Bucy Erik P

Cabri G

Cacciapuoti Angela Sara

Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao

Christodoulakis Stavros

Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin

Gudmundsen Magne Guido MarkHanna Alex

Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y

Karanastasi Anastasia

Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell

Pierre SamuelReichenbacher Tumasch

Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun

Yeongkiloon Chang-Yoon Won-Jung

Figure 11 Cooperation network of 65 authors (node colors represent different countriesregions eg range for South Korea red for the USApurple for Australia green for China and brown for Italy)The network can be accessed via the link (httpwwwzhukunorghaotyresourcesaspid=NLPEMC coauthor)

0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)

Figure 12 (a) Estimated120572 value for themodels fitted usingVEM (b) Perplexities of the test data for themodels fitted by usingGibbs samplingEach line corresponded to one of the folds in the 10-fold cross-validation


Topic_39Topic_38Topic_37Topic_35Topic_34Topic_33Topic_32Topic_31Topic_30Topic_29Topic_28Topic_27Topic_26Topic_25Topic_24Topic_23Topic_22Topic_21Topic_20Topic_19Topic_18Topic_16Topic_15Topic_14Topic_13Topic_12Topic_10Topic_9Topic_8Topic_7Topic_6Topic_3Topic_2Topic_5Topic_17Topic_40Topic_1Topic_4Topic_36Topic_11

Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11

Figure 13 The visualized result of hierarchical clustering based on term-topic posterior probability matrix

4 Discussions

This study provides a most up-to-date bibliometric analysison the publications in WoS during the years 2000ndash2016 inthe NLP empowered mobile computing research field Someinteresting findings are discussed below

The annual number of the publication distribution showsa significant growth trend from 12 publications in 2000 to 55publications in 2016 This indicates a growing interest in theresearch field

The literature characteristics analysis shows that the 417publications are widely dispersed throughout 287 journals11 most productive journals together contribute about 21 ofthe total publications The top 3 are IEEEACM Transactions

on Audio Speech and Language Processing Speech Communi-cation andComputer Speech and LanguageComputer scienceis the most shared subject among these 11 journals JournalInformation Sciences possesses the highest IF SJR 5-Year IFand CiteScore except for the SNIP score in year 2016

Top 3 most influential publications are [35] by Miaoet al published in 2010 [36] by MacKenzie and Soukoreffpublished in 2002 and [37] by Strayer and Drews publishedin 2007

There are 1408 authors and 544 affiliations involved in thepublications Most authors (7918) have only 1 publicationand 425 of the authors have 3 or more publications Themost productive authors are Chen Tao from Singapore andMizzaro Stefano from Italy In addition most affiliations


2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16

Topic 17 Topic 18 Topic 19 Topic 20 Topic 21

00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04

Topic 25 Topic 26 Topic 27 Topic 28 Topic 29 Topic 30 Topic 31 Topic 32


00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08

Figure 14 The trends of the 40 research topics during 2000ndash2016 (x-coordinate as year y-coordinate as proportion )

Table 8 Top 20 most frequent terms

Rank Stemmed terms Occurrence numberTotal 2000ndash2008 2009ndash2016

1 Agent 369 250 1192 Image 215 70 1453 Sentiment 128 0 1284 Dialogue 83 49 345 Health 81 2 796 Music 76 27 497 Radio 74 10 648 Unit 74 51 239 Adaptation 70 40 3010 Relevance 69 29 4011 Geographic 66 37 2912 Short Messages 66 9 5713 Protocol 65 20 4514 Chinese 64 29 3515 Medical 60 16 4416 Recommendation 60 4 5617 Clustering 54 20 3418 Privacy 54 9 4519 Ad hoc 53 9 4420 Traffic 52 17 35

(7006) have 1 publication 1189 of the affiliations have3 or more publications The most productive affiliationsare Nanyang Technological University from Singapore andTsinghua University from China Lee Chin-Hui from USAwith 5767 ACP ranks 1st among top 20 productive authorsand Georgia Institute of Technology from USA with 110 ACPranks 1st among 15 most productive affiliations

Through geographic visualization analysis 60 coun-triesregions have participated in the publications Thetop 15 productive countriesregions are developed coun-triesregions except for China As the top 2 the USA andChina have shown a significant growth in the numbers of sci-entific publications since 2010 These numbers are predictedto continue to increase in the coming years This partiallyreflects the need of the development of NLP techniques insolving mobile computing issues

Scientific collaboration analysis shows that there are sig-nificant growth of international collaborations institution-collaborations as well as author-collaborations Throughsocial network analysis we found that researchers tend tocollaborate with others within the same country or areawith institutions under similar administration or with aneighboring country or area However some research institu-tions might have separate administration arrangements fromtheir associated universities or hospitals and a researchermight be affiliated with multiple institutions The co-authors


Table 9 Top 15 most frequent terms for the top 10 best matching topics

Topic Potential theme Top high frequency terms

36 Mobile agent computing Agent Coordination Java Migration Protocol Mobile-agent Failure ItineraryFiltering Turkish Attack Commerce Context-aware Truncation Crash

11 Mobile agent computingAgent Planning Ontology Cloud Multi-agent Net Interoperability Neural

Peer-to-Peer Broadband Instruction Complementarity Natural Language TrafficGrounding

32 Mobile privacy and securityPrivacy Private Secure Location-Based Services Encryption Points of InterestProtection Approximate Attack Path Privacy-preserving Streaming Password

Protocol Cryptosystem

1 Image and syllable eventsImage Particular Allophones Re-ranking Composite Phoneme Simple PhonemesSyllable Thing iPad On-Premise Signs Spreading Bow Modern Orthography

Arabic Content-based Descriptor

4 Mobile social media computing Sentiment Opinion Twitter Tweet Customer Suggestion Emojis EmotionMicro-blog Protest Brand Suggestive Microblog Orientation Box

8 Mobile radio Radio Phone-in Localization Australian Formulation Island ReportingTalkback Involvement Caller Dialogic Stance Backlinking Cloud French

5 Mobile location computingGeographic Relevance Seeking Innovation Subspace Tourism Birthright FirmFlier Sensing TILES (Temporal Identity Location Environmental and Social)

Cross-space Location-aware Personalized Reposting

40 Context-aware computingDialogue Context-aware Estonian Clarification Array Problematic

Reformulation Verbose Email Mobile Information Services enabled by MobilePublishing Non-understanding Publishing Agent Directive Reinforcement

10 Second screen response Gesture Debate PreFrontal Cortex Adult Presidential Walking Facial TwitterEducational Gait Political Touch Biometrics Blink Cortex

35 Language learning and modelingChinese Information Retrieval Peer-to-Peer Conditional Random Field UpdateApprentice Affordances Disyllabic Website Workplace Self-study Skip-chain

Descriptive Mobile Peer-to-Peer Multilingual

might actually work together but are affiliated with differentinstitutions Therefore it is worth noticing that institution-wise collaboration might not be the actual collaborationamong institutions

Most topics identified using LDA method are recogniz-able as they are related to major issues in the research fieldDue to space constraints herewe only provide interpretationsof some representative topics

Topic 36 and Topic 11 contain words such as ldquoAgentrdquoldquoMobile-agentrdquo ldquoMulti-agentrdquo ldquoItineraryrdquo ldquoMigrationrdquoldquoProtocolrdquo and ldquoTruncationrdquo Thus Topic 36 and Topic11 pertain to mobile agent computing As an emerging andexciting paradigm for mobile computing applications [42]mobile agent can not only support mobile computers anddisconnected operations but also provide an efficientconvenient and robust programming paradigm forimplementing distributed applications The use of mobileagent can bring about significant benefits eg reduction ofnetwork traffic overcoming network latency and seamlesssystem integration Therefore mobile agent is well adaptedto the domain of mobile computing

Topic 32 discusses events about mobile privacy andsecurity Words in this topic include ldquoPrivacyrdquo ldquoPrivaterdquoldquoSecurerdquo ldquoEncryptionrdquo ldquoPrivacy-preservingrdquo ldquoPasswordrdquoand ldquoCryptosystemrdquo As pointed out by Mollah et al [43]security andprivacy challenges are introducedwith the devel-opment of mobile cloud computing which aims at relievingchallenges of the resource constrained mobile devices in

mobile computing area Studies centering on mobile privacycan be found For example Xi et al [44] applied PrivateInformation Retrieval techniques in finding the shortest pathbetween an origin and a destination in location privacy issueswithout the risk of disclosing their privacy

Topic 1 discusses mobile computing on image and syllableevents It includes words such as ldquoImagerdquo ldquoSyllablerdquo ldquoRe-rankingrdquo ldquoContent-basedrdquo ldquoComposite Phonemerdquo ldquoSimplePhonemesrdquo and ldquoModern Orthographyrdquo Image search inmobile device is quite worthy of challenge [45] Manyresearchers are seeking ways to solve this problem Forexample Cai et al [46] presented a new geometric rerankingalgorithm specific for small vocabulary in aforementionedscenarios based on Bag-of-Words model for image retrievalMobile computing on syllable events is another focus Arepresentative work is by Eddington and Elzinga [47] Theyconducted a quantitative analysis on the phonetic contextof word-internal flapping with great attention paid to stressplacement following phone and syllabification

Topic 4 mainly focuses on mobile social media eventWords like ldquoTwitterrdquo ldquoSentimentrdquo ldquoTweetrdquo ldquoEmojisrdquoldquoMicro-blogrdquo ldquoOpinionrdquo ldquoPublicrdquo and ldquoEmotionrdquo can befound within this topic With the rapid development ofsocial network information spreading and evolution isfacilitated with popularity of the environment of wirelesscommunication especially social media platform on mobileterminals [48] Researchers are gradually paying attentionto this area For example based on 100 million collected


messages from Twitter Wang et al [49] presented a hybridmodel for sentimental entity

Based on topic distributions we found that mobile agentcomputing mobile social media computing and sound relatedevent computing are 3 highest-frequent research themesFromFigure 14 aswell asMannndashKendall test results we foundthat some research themes present a statistically significantincreasing trend eg image and syllable related eventsmobilesocial media computing and healthy related events whileresearches on mobile agent computing presents a statisticallysignificant decreasing trend

In the thematic analysis the optimal number of topicswasselected as 40 by a statistical measure of model fitting thedata However mechanical reliance on statistical measuresmight lead to the selection of a less meaningful topic model[50] Hence we manually checked the robustness of theresults by confirming identified topics using a qualitativeassessment with the basis of prior knowledge For each topicwe checked the semantic coherence of its high-frequencyterms and examined the contents of publication with a highproportion of this topic

Through the AP clustering analysis on the 40-topics 8clusters were identified ie mobile agent computing mobilesocial media computing image and syllable related eventscontext-aware computing sound related events mobile loca-tion computing healthy related events and other events Theresults of AP clustering analysis are on the whole sensibleand easy-to-understand However we still found that the8 categories vary a lot in topic numbers One possiblereason is the choice of clustering method We then adoptedhierarchical clustering method with category number settingto 8 The result was similar with AP clustering Anotherpossible reason is the sample size since the number of therelevant publications in WoS is limited

This study is the first to thoroughly explore researchstatus of the NLP empowered mobile computing researchfield in the statistical perspective The study provides acomprehensive overview and an intellectual structure of thefield from 2000 to 2016 The findings can potentially helpresearchers especially newcomers systematically understandthe development of the field learn the most influentialjournals recognize potentially academic collaborators andtrace research hotspots

For future work there are several directions First morecomprehensive data is expected to be included ThoughWoSis a widely applied repository for bibliometric analysis due toits high authority some relevant conference proceedings havenot been indexed yet in WoS Second we intend to employdifferent data clustering methods and compare clusteringresults for deeper cluster analyzing

5 Conclusions

We conducted a bibliometric analysis on natural languageprocessing empowered mobile computing research publica-tions fromWeb of Science published during years 2000ndash2016The literature characteristics were uncovered using a descrip-tive statistics method Geographical publication distributionwas explored using a geographic visualization method By

applying a social network analysis method cooperation rela-tionships among countriesregions affiliations and authorswere displayed Finally topic discovery and distribution werepresented using a LDAmethod and anAP clusteringmethodWe believe the analysis can help researchers comprehend thecollaboration patterns and distribution of scholarly resourcesand research hot spots in the research field more systemati-cally

Disclosure

Tianyong Hao and Yi Zhou are the corresponding authors

Conflicts of Interest

The authors declare that they have no conflicts of interest

Acknowledgments

The work was substantially supported by the grant fromNational Natural Science Foundation of China (no61772146) the Innovative School Project inHigher Educationof Guangdong Province (No YQ2015062) Science andTechnology Program of Guangzhou (no 201604016136) andMajor Project of Frontier and Key Technical Innovation ofGuangdong Province (no 2014B010118003)

References

[1] G Deepak and B S Pradeep ldquoChallenging issues and limita-tions of mobile computingrdquo vol 3 pp 177ndash181 2012

[2] K-Y Chung J Yoo and K J Kim ldquoRecent trends on mobilecomputing and future networksrdquo Personal and UbiquitousComputing vol 18 no 3 pp 489ndash491 2014

[3] M Chen J Pan Q Zhao and Y Yan ldquoMulti-task learningin deep neural networks for Mandarin-english code-mixingspeech recognitionrdquo IEICE Transaction on Information andSystems vol E99D no 10 pp 2554ndash2557 2016

[4] N Ilayaraja F Mary Magdalene Jane M Safar and R Nadara-jan ldquoWARM Based Data Pre-fetching and Cache ReplacementStrategies for LocationDependent Information System inWire-less Environmentrdquo Wireless Personal Communications vol 90no 4 pp 1811ndash1842 2016

[5] L-H Wong R B King C S Chai and M Liu ldquoSeamlesslylearning Chinese contextual meaning making and vocabularygrowth in a seamless Chinese as a second language learningenvironmentrdquo Instructional Science vol 44 no 5 pp 399ndash4222016

[6] O J Rasanen and J P Saarinen ldquoSequence prediction withsparse distributed hyperdimensional coding applied to theanalysis of mobile phone use patternsrdquo IEEE Transactions onNeural Networks and Learning Systems vol 27 no 9 pp 1878ndash1889 2016

[7] M Puppala T He S Chen et al ldquoMETEOR An EnterpriseHealth Informatics Environment to Support Evidence-BasedMedicinerdquo IEEE Transactions on Biomedical Engineering vol62 no 12 pp 2776ndash2786 2015

[8] A O Adesina K K Agbele A P Abidoye andH O NyongesaldquoText messaging and retrieval techniques for a mobile health


information systemrdquo Journal of Information Science vol 40 no6 pp 736ndash748 2014

[9] W-T Chiu and Y-S Ho ldquoBibliometric analysis of tsunamiresearchrdquo Scientometrics vol 73 no 1 pp 3ndash17 2007

[10] J M Merigo A M Gil-Lafuente and R R Yager ldquoAn overviewof fuzzy research with bibliometric indicatorsrdquo Applied SoftComputing vol 27 pp 420ndash433 2015

[11] D Bouyssou and T Marchant ldquoRanking scientists and depart-ments in a consistent mannerrdquo Journal of the Association forInformation Science and Technology vol 62 no 9 pp 1761ndash17692011

[12] A Mazloumian ldquoPredicting Scholarsrsquo Scientific Impactrdquo PLoSONE vol 7 no 11 Article ID e49246 2012

[13] X Chen H Xie F Wang Z Liu J Xu and T Hao ldquoNaturalLanguage Processing in Medical Research A BibliometricAnalysisrdquo BMC Medical Informatics and Decision Making vol18 supplement 1 no 14 2018

[14] Y Geng W Chen Z Liu et al ldquoA bibliometric review Energyconsumption and greenhouse gas emissions in the residentialsectorrdquo Journal of Cleaner Production vol 159 pp 301ndash316 2017

[15] A Khan N Choudhury S Uddin L Hossain and L ABaur ldquoLongitudinal trends in global obesity research andcollaboration A review using bibliometric metadatardquo ObesityReviews vol 17 no 4 pp 377ndash385 2016

[16] N Roig-Tierno T F Gonzalez-Cruz and J Llopis-MartinezldquoAnoverviewof qualitative comparative analysis A bibliometricanalysisrdquo Journal of Innovation Knowledge vol 2 no 1 pp 15ndash23 2017

[17] G Albort-Morant and D Ribeiro-Soriano ldquoA bibliometricanalysis of international impact of business incubatorsrdquo Journalof Business Research vol 69 no 5 pp 1775ndash1779 2016

[18] J M Merigo and J-B Yang ldquoA bibliometric analysis ofoperations research and management sciencerdquo OMEGA - TheInternational Journal of Management Science vol 73 pp 37ndash482017

[19] K Zhang Q Wang Q-M Liang and H Chen ldquoA bibliometricanalysis of research on carbon tax from 1989 to 2014rdquoRenewableamp Sustainable Energy Reviews vol 58 pp 297ndash310 2016

[20] K Randhawa R Wilden and J Hohberger ldquoA BibliometricReview of Open Innovation Setting a Research AgendardquoJournal of Product Innovation Management vol 33 no 6 pp750ndash772 2016

[21] A Yataganbaba and I Kurtbas ldquoA scientific approach withbibliometric analysis related to brick and tile drying A reviewrdquoRenewable amp Sustainable Energy Reviews vol 59 pp 206ndash2242016

[22] X Chen B Chen C Zhang and T Hao ldquoDiscovering theRecent Research inNatural Language Processing Field Based ona Statistical Approachrdquo in Emerging Technologies for Educationvol 10676 of Lecture Notes in Computer Science pp 507ndash517Springer International Publishing Cham 2017

[23] H J Kim D Y Yoon E S Kim K Lee J S Bae and J-H LeeldquoThe 100 most-cited articles in neuroimaging A bibliometricanalysisrdquo Results in Physics vol 139 pp 149ndash156 2016

[24] X Chen H Weng and T Hao ldquoA Data-Driven Approachfor Discovering the Recent Research Status of Diabetes inChinardquo Lecture Notes in Computer Science (including subseriesLecture Notes in Artificial Intelligence and Lecture Notes inBioinformatics) Preface vol 10594 pp 89ndash101 2017

[25] W R Tobler ldquoA computer movie simulating urban growth inthe detroit regionrdquo Economic Geography vol 46 supplement 1pp 234ndash240 1970

[26] E Otte and R Rousseau ldquoSocial network analysis a powerfulstrategy also for the information sciencesrdquo Journal of Informa-tion Science vol 28 no 6 pp 441ndash453 2002

[27] D R Radev M T Joseph B Gibson and P Muthukrishnan ldquoAbibliometric and network analysis of the field of computationallinguisticsrdquo Journal of the Association for Information Scienceand Technology vol 67 no 3 pp 683ndash706 2016

[28] M Grandjean ldquoA social network analysis of Twitter Mappingthe digital humanities communityrdquo Cogent Arts and Humani-ties vol 3 no 1 Article ID 1171458 2016

[29] M Salathe M Kazandjieva J W Lee P Levis M W Feldmanand J H Jones ldquoA high-resolution human contact networkfor infectious disease transmissionrdquo Proceedings of the NationalAcadamy of Sciences of the United States of America vol 107 no51 pp 22020ndash22025 2010

[30] J Scott ldquoSocial network analysisrdquo Sage 2017[31] D M Blei A Y Ng and M I Jordan ldquoLatent Dirichlet

allocationrdquo Journal ofMachine Learning Research vol 3 no 4-5pp 993ndash1022 2003

[32] J R Finkel T Grenager and C Manning ldquoIncorporating non-local information into information extraction systems by Gibbssamplingrdquo in Proceedings of the the 43rd Annual Meeting pp363ndash370 Ann Arbor Michigan June 2005

[33] B J Frey and D Dueck ldquoClustering by passing messagesbetween data pointsrdquo American Association for the Advance-ment of Science Science vol 315 no 5814 pp 972ndash976 2007

[34] A F El-Samak and W Ashour ldquoOptimization of TravelingSalesman Problem Using Affinity Propagation Clustering andGenetic Algorithmrdquo Journal of Artificial Intelligence and SoftComputing Research vol 5 no 4 2015

[35] G Miao N Himayat and G Y Li ldquoEnergy-efficient linkadaptation in frequency-selective channelsrdquo IEEE Transactionson Communications vol 58 no 2 pp 545ndash554 2010

[36] I S MacKenzie and R W Soukoreff ldquoText entry formobile computing Models and methods theory and practicerdquoHumanndashComputer Interaction vol 17 no 2-3 pp 147ndash198 2002

[37] D L Strayer and F A Drews ldquoCell-phone-induced driverdistractionrdquo Current Directions in Psychological Science vol 16no 3 pp 128ndash131 2007

[38] J Cao T Chen and J Fan ldquoLandmark recognition withcompact BoWhistogramand ensemble ELMrdquoMultimediaToolsand Applications 2015

[39] MMMostafa ldquoMore than words social networksrsquo text miningfor consumer brand sentimentsrdquo Expert Systems with Applica-tions vol 40 no 10 pp 4241ndash4251 2013

[40] H Jiang M Qiang and P Lin ldquoA topic modeling basedbibliometric exploration of hydropower researchrdquo Renewable ampSustainable Energy Reviews vol 57 pp 226ndash237 2016

[41] H BMann ldquoNonparametric tests against trendrdquo Econometricavol 13 pp 245ndash259 1945

[42] D B Lange and M Oshima ldquoSeven Good Reasons for MobileAgentsrdquo Communications of the ACM vol 42 no 3 pp 88-891999

[43] M B Mollah M A K Azad and A Vasilakos ldquoSecurity andprivacy challenges in mobile cloud computing Survey and wayaheadrdquo Journal of Network and Computer Applications vol 84pp 38ndash54 2017

[44] Y Xi L Schwiebert and W Shi ldquoPrivacy preserving shortestpath routing with an application to navigationrdquo Pervasive andMobile Computing vol 13 pp 142ndash149 2014


[45] T Yan V Kumar and D Ganesan ldquoCrowdSearch exploitingcrowds for accurate real-time image search on mobile phonesrdquoin Proceedings of the 8th International Conference on MobileSystems Applications and Services (MobiSys rsquo10) pp 77ndash90ACM San Francisco Calif USA June 2010

[46] Y Cai S Li Y Cheng and R Ji ldquoLocal consistent hierarchicalHough Match for image re-rankingrdquo Journal of Visual Commu-nication and Image Representation vol 37 pp 32ndash39 2016

[47] D Eddington and D Elzinga ldquoThe phonetic context of amer-ican english flapping Quantitative evidencerdquo Language andSpeech vol 51 no 3 pp 245ndash266 2008

[48] X Wang H Zhang S Yuan J Wang and Y Zhou ldquoSentimentprocessing of social media information from both wireless andwired networkrdquo EURASIP Journal on Wireless Communicationsand Networking vol 2016 no 1 article no 164 2016

[49] Z Wang X Cui L Gao Q Yin L Ke and S Zhang ldquoAhybridmodel of sentimental entity recognition onmobile socialmediardquo EURASIP Journal on Wireless Communications andNetworking vol 2016 no 1 article no 253 2016

[50] K E C Levy andM Franklin ldquoDrivingRegulationUsingTopicModels to Examine Political Contention in the US TruckingIndustryrdquo Social Science Computer Review vol 32 no 2 pp182ndash194 2014

International Journal of

AerospaceEngineeringHindawiwwwhindawicom Volume 2018

RoboticsJournal of

Hindawiwwwhindawicom Volume 2018


Active and Passive Electronic Components

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawiwwwhindawicom

Volume 2018

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018

Control Scienceand Engineering

Journal of



Journal ofEngineeringVolume 2018

SensorsJournal of



RotatingMachinery


Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation


Hindawi

wwwhindawicom Volume 2018

Advances in

Multimedia

Submit your manuscripts atwwwhindawicom


information to networks and to improve the prediction oflanguage switching for the primary task of senone classifi-cation The proposed schemes enhanced the recognition onboth languages and reduced the relative overall error ratesby 44 on average when dealing with real-world Mandarin-English corpus in mobile voice search Ilayaraja et al [4]presented a weighted association rule mining prefetchingtechnique to determine the secondary service item withthe consideration of access frequency of services semanticdistance among the successive query request and spatialdistance between service instances and user context (egposition service type and query request time) Wong et al[5] analyzed the studentsrsquo vocabulary usage using a corpusanalysis tool to identify and unpack the contextual conditionsin which a mobile- and cloud-assisted Chinese languagelearning environment promoted key learning outcomesRasanen and Saarinen [6] proposed a method based onsparse hyperdimensional coding of sequence structures forsequence prediction Their experiments suggested that themethod was capable of capturing the relevant variable-orderstructure from the sequences ANLP based toolMOTTEwasdeveloped by Puppala et al [7] for extracting and structuringdata in pathology reports automatically to support clinicalsolution applications With an aim of screening informationon human immunodeficiency virusacquired immune defi-ciency syndrome Adesina et al [8] designed a monolingualshort message services based system for the retrieval offrequently asked questions

Bibliometric analysis is defined as the use of statisti-cal methods on evaluating scholarly publications from anobjective and quantitative perspective within a certain field[9] Benefits of bibliometric analysis include (1) organizinginformation in a specific thematic field [10] (2) evaluatingscientific developments in knowledge of a specific subjectand assessing the scientific quality [11] (3) determining theimpact of research funding (4) comparing research perfor-mance across different affiliations and document changes inthe research workforce and (5) identifying emerging areasof research focus and predicting future research success[12] As for researchers especially newcomers bibliometricanalysis can assist them in (1) better selecting potentialresearch topics (2) demonstrating the values and impacts oftheir relevant works (3) recognizing appropriate academicresearchers to seek research collaboration and (4) keepingabreast of new research status and new technological changes[13]

Bibliometric analysis has been widely applied to variousfields for the measurement of quality and productivity ofacademic output and has demonstrated excellent effective-ness from long-term practice Relevant researches mainlyfocused on revealing publication statistical characteristicsexploring the collaboration relationship and uncoveringresearch themes and their evolution Some examples are asfollows Geng et al [14] conducted a bibliometric survey ofthe research field of residential energy and greenhouse gasemissions for the purpose of uncovering research status Intheir work citation analysis was used to assess the influenceof journals countries and authors while network analysiswas performed to evaluate the relationships among countries

authors and keywords Based on 117340 obesity-relatedresearch publications indexed in Scopus database publishedfrom 1993ndash2012 Khan et al [15] reported research trendsand collaboration patterns in the field Roig-Tierno et al[16] conducted a bibliometric analysis on research publica-tions with the application of qualitative comparative analysis(QCA) Their study revealed the differences in quantitativeterms of the three variants of QCA Albort-Morant andRibeiro-Soriano [17] focused on the research developmentof business incubators They sorted 445 publications fromWoS according to bibliographic indicators such as researcharea and year of publication Their study revealed the lackof publications on business incubators and highlighted thefragmented nature of research themes Merigo and Yang[18] aimed at identifying relevant researches and the newesttrends in field of operation research andmanagement scienceThe analysis involved some influential journals two hun-dred most cited publications and productive and influentialauthors Zhang et al [19] quantitatively and qualitativelyevaluated carbon tax related literature from 1989 to 2014using bibliometric analysis Their study demonstrated thatthe USA was the leading country and the Vrije UniversityAmsterdam and Massachusetts Institute of Technology andStanford University were the most productive affiliations inthe research field Randhawa et al [20] conducted a system-atic review of publications on open innovation (OI) researcharea using bibliometrics cocitation analysis and text miningThree distinct areas within OI research were identified iefirm-centric aspects of OI management of OI networks androle of users and communities in OI In order to discover theworldwide trends in the research field of drying bricktileYataganbaba and Kurtbas [21] analyzed relevant patents interms of eg publication number authorship and ownershipand international collaboration patterns Merigo et al [10]explored the research development trends in fuzzy sciencesSimilar works have also been conducted in other fields egnatural language processing [22] neuroimaging [23] anddiabetes [24]

To the best of our knowledge there is no scientificreview of NLP empowered mobile computing research fieldcurrently Thus in this study we conduct a bibliometricanalysis on publications retrieved fromWoS during the years2000ndash2016 to explore the research status of the research fieldThe main objective is to address the following issues (1)investigating publication statistical characteristics and publi-cation collaborations (2) exploring publication geographicaldistributions (3) visualizing scientific collaboration relation-ships and (4) reveling current hot research topic themes andresearch topic changes

The rest of the paper is organized as follows Section 2introduces methods and materials Bibliometric analysisresults on retrieved research publications are reported inSection 3 Findings and discussion are shown in Section 4while Section 5 summarizes the work

2 Methods and Materials

Five different methods are applied to analyze research pub-lications in the NLP empowered mobile computing field



21 Methods
















ℓ (120572 120573) = log (119901 (119889 | 120572 120573))


= logintsum119911

[ 119873prod119894=1

119901 (119908119894 | 119911119894 120573) 119901 (119911119894 | 120579)]sdot 119901 (120579 | 120572) 119889120579

(1)


log (119901 (119889 | 119911)) = 119896 log(Γ (119881120575)Γ (120575)119881)

+ 119896sum119870=1

[[119881sum119895=1


(2)



(3)


log (119901 (119889)) = 119863sum119889=1

119881sum119895=1

119899(119895119889) log[ 119896sum119870=1

120579(119889)119870 120573(119895)119870 ] (4)

















119903 (119894 119895) larr997888 119904 (119894 119895) minus max1198951015840st 1198951015840 =119895

119886 (119894 1198951015840) + 119904 (119894 1198951015840)

119886 (119894 119895)119894 =119895

larr997888 min0 119903 (119894 119895) + sum

forall1198941015840notin119894119895

max 0 119903(1198941015840 119895119886 (119895 119895) larr997888 sum

1198941015840st 1198941015840 =119895max 0 119903 (1198941015840 119895)

(6)


119862119867119894 = argmax119895119886 (119894 119895) + 119903 (119894 119895) (7)







3 Results














(3270) Retrieval (3142)




(a) (b)

(c) (d)


055110165220275330385440495550

0153045607590

105120135150

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

cita

tion

pub

av

g

cita

tion

per p

ub

and

ann

ual c

itatio

n

pubcitation




Speech Commun

Comput Speech Lang


Expert Syst Appl

IEEE

Electron

Trans Consum

Mob Inf Syst

Multimed Tools Appl



Commun


2491081325013143

35 238 267 47 256 103 119 312 537 136 049

IEICE Trans Inf

Syst

0

05

1

15

2

25

3

35

4

45

5
















Son D J et al 2004 77 592























lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia




21 Methods
















ℓ (120572 120573) = log (119901 (119889 | 120572 120573))


= logintsum119911

[ 119873prod119894=1

119901 (119908119894 | 119911119894 120573) 119901 (119911119894 | 120579)]sdot 119901 (120579 | 120572) 119889120579

(1)


log (119901 (119889 | 119911)) = 119896 log(Γ (119881120575)Γ (120575)119881)

+ 119896sum119870=1

[[119881sum119895=1


(2)



(3)


log (119901 (119889)) = 119863sum119889=1

119881sum119895=1

119899(119895119889) log[ 119896sum119870=1

120579(119889)119870 120573(119895)119870 ] (4)

















119903 (119894 119895) larr997888 119904 (119894 119895) minus max1198951015840st 1198951015840 =119895

119886 (119894 1198951015840) + 119904 (119894 1198951015840)

119886 (119894 119895)119894 =119895

larr997888 min0 119903 (119894 119895) + sum

forall1198941015840notin119894119895

max 0 119903(1198941015840 119895119886 (119895 119895) larr997888 sum

1198941015840st 1198941015840 =119895max 0 119903 (1198941015840 119895)

(6)


119862119867119894 = argmax119895119886 (119894 119895) + 119903 (119894 119895) (7)







3 Results














(3270) Retrieval (3142)




(a) (b)

(c) (d)


055110165220275330385440495550

0153045607590

105120135150

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

cita

tion

pub

av

g

cita

tion

per p

ub

and

ann

ual c

itatio

n

pubcitation




Speech Commun

Comput Speech Lang


Expert Syst Appl

IEEE

Electron

Trans Consum

Mob Inf Syst

Multimed Tools Appl



Commun


2491081325013143

35 238 267 47 256 103 119 312 537 136 049

IEICE Trans Inf

Syst

0

05

1

15

2

25

3

35

4

45

5
















Son D J et al 2004 77 592























lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



= logintsum119911

[ 119873prod119894=1

119901 (119908119894 | 119911119894 120573) 119901 (119911119894 | 120579)]sdot 119901 (120579 | 120572) 119889120579

(1)


log (119901 (119889 | 119911)) = 119896 log(Γ (119881120575)Γ (120575)119881)

+ 119896sum119870=1

[[119881sum119895=1


(2)



(3)


log (119901 (119889)) = 119863sum119889=1

119881sum119895=1

119899(119895119889) log[ 119896sum119870=1

120579(119889)119870 120573(119895)119870 ] (4)

















119903 (119894 119895) larr997888 119904 (119894 119895) minus max1198951015840st 1198951015840 =119895

119886 (119894 1198951015840) + 119904 (119894 1198951015840)

119886 (119894 119895)119894 =119895

larr997888 min0 119903 (119894 119895) + sum

forall1198941015840notin119894119895

max 0 119903(1198941015840 119895119886 (119895 119895) larr997888 sum

1198941015840st 1198941015840 =119895max 0 119903 (1198941015840 119895)

(6)


119862119867119894 = argmax119895119886 (119894 119895) + 119903 (119894 119895) (7)







3 Results














(3270) Retrieval (3142)




(a) (b)

(c) (d)


055110165220275330385440495550

0153045607590

105120135150

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

cita

tion

pub

av

g

cita

tion

per p

ub

and

ann

ual c

itatio

n

pubcitation




Speech Commun

Comput Speech Lang


Expert Syst Appl

IEEE

Electron

Trans Consum

Mob Inf Syst

Multimed Tools Appl



Commun


2491081325013143

35 238 267 47 256 103 119 312 537 136 049

IEICE Trans Inf

Syst

0

05

1

15

2

25

3

35

4

45

5
















Son D J et al 2004 77 592























lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia






119903 (119894 119895) larr997888 119904 (119894 119895) minus max1198951015840st 1198951015840 =119895

119886 (119894 1198951015840) + 119904 (119894 1198951015840)

119886 (119894 119895)119894 =119895

larr997888 min0 119903 (119894 119895) + sum

forall1198941015840notin119894119895

max 0 119903(1198941015840 119895119886 (119895 119895) larr997888 sum

1198941015840st 1198941015840 =119895max 0 119903 (1198941015840 119895)

(6)


119862119867119894 = argmax119895119886 (119894 119895) + 119903 (119894 119895) (7)







3 Results














(3270) Retrieval (3142)




(a) (b)

(c) (d)


055110165220275330385440495550

0153045607590

105120135150

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

cita

tion

pub

av

g

cita

tion

per p

ub

and

ann

ual c

itatio

n

pubcitation




Speech Commun

Comput Speech Lang


Expert Syst Appl

IEEE

Electron

Trans Consum

Mob Inf Syst

Multimed Tools Appl



Commun


2491081325013143

35 238 267 47 256 103 119 312 537 136 049

IEICE Trans Inf

Syst

0

05

1

15

2

25

3

35

4

45

5
















Son D J et al 2004 77 592























lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia













(3270) Retrieval (3142)




(a) (b)

(c) (d)


055110165220275330385440495550

0153045607590

105120135150

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

cita

tion

pub

av

g

cita

tion

per p

ub

and

ann

ual c

itatio

n

pubcitation




Speech Commun

Comput Speech Lang


Expert Syst Appl

IEEE

Electron

Trans Consum

Mob Inf Syst

Multimed Tools Appl



Commun


2491081325013143

35 238 267 47 256 103 119 312 537 136 049

IEICE Trans Inf

Syst

0

05

1

15

2

25

3

35

4

45

5
















Son D J et al 2004 77 592























lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



(a) (b)

(c) (d)


055110165220275330385440495550

0153045607590

105120135150

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

cita

tion

pub

av

g

cita

tion

per p

ub

and

ann

ual c

itatio

n

pubcitation




Speech Commun

Comput Speech Lang


Expert Syst Appl

IEEE

Electron

Trans Consum

Mob Inf Syst

Multimed Tools Appl



Commun


2491081325013143

35 238 267 47 256 103 119 312 537 136 049

IEICE Trans Inf

Syst

0

05

1

15

2

25

3

35

4

45

5
















Son D J et al 2004 77 592























lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



Speech Commun

Comput Speech Lang


Expert Syst Appl

IEEE

Electron

Trans Consum

Mob Inf Syst

Multimed Tools Appl



Commun


2491081325013143

35 238 267 47 256 103 119 312 537 136 049

IEICE Trans Inf

Syst

0

05

1

15

2

25

3

35

4

45

5
















Son D J et al 2004 77 592























lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia











Son D J et al 2004 77 592























lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia













lt2000

[2000 11000)[11000 17000)

[17000 37000](37000 61000](61000 105000]

gt105000









2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia




2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

5

10

15

20

in

tern

atio

nal-c

olla

bora

tive p

ub

0

5

10

15

20

25

30

35

Perc

enta

ge o

f int

erna

tiona

l-co

llabo

rativ

e pub

(

)


2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


0

10

20

30

40

50

60

70

Perc

enta

ge o

f ins

titut

ion-

colla

bora

tive p

ub

0

5

10

15

20

25

30

35

in

stitu

tion-

colla

bora

tive p

ub




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia




USA

Peoples R China

EnglandAustralia

France

Hong Kong

Italy

Singapore

Switzerland

Finland

Germany

Japan

Saudi Arabia

South Korea

Taiwan

Canada

SpainMalaysia

Bulgaria

Malta

India

LuxembourgNorway

Pakistan

South Africa

TurkeyIreland

Jordan

Netherlands

Vietnam

Algeria

Argentina

Austria


Colombia

Croatia

Denmark

Ethiopia

Georgia

Israel

Kuwait

LithuaniaMexico

Serbia

Slovenia

Sweden

Thailand






Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



Aalto Univ

ASTAR

Beihang Univ



China Univ Geosci


Chungnam Natl Univ

City Univ Hong Kong

Clemson Univ


Dalian Univ Technol

Fdn Ugo Bordoni



Haukeland Hosp


Inha Univ



Intel Corp

Kangwon Natl Univ


King Saud Univ


Kwangwoon Univ

Kyungpook Natl U


Microsoft Res

Middle E Tech Univ

Mitre Corp

Nanyang Technol

Natl Chengchi Univ




Natl Taipei Univ




Natl Univ Singapore

North West Univ



Oakland Univ


Penn State Univ

Purdue Univ


Renmin Univ China

Sogang Univ

SRI Int



UCL

Univ Athens

Univ

Univ Calif Berkeley


Univ Colorado

Univ Granada

Univ Hong Ko

Univ Illinois

Univ Malaya

Univ Newcastle

Univ Oxford

Univ S Carolina


Univ Sheffield


Univ Thessaly

Univ Tokyo

Univ Toronto

Univ Udine

Univ Washington


Univ Zurich

Vanderbilt Univ

Victoria Univ

Wuhan Univ

Xiamen Univ

Yeungn

York Univ

Zhejiang Univ







Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



Alkhairy Ashraf

Baek Jin-Wook


Bless Josef J

Bucy Erik P

Cabri G


Caleffi Marcello

Carpineto Claudio

Chen Chao

Chen Shu-Ching

Chen Tao


Davel Marelie

De Sabbata Stefano

Delot T

Eggel Ivan

Furui Sadaoki

Grover Justin


Hiyama Aya

Hugdahl Kenneth

Hurson AR

Hwang Mei-Yuh

Jafri Afshan

Jiao Y


Kazasis Fotis G

Kim Harksoo

Kompus Kristiina

Leonardi LLi Ming

Lu Ping

Mizzaro Stefano

Moon Seok-Jae

Mueller Henning

Muneyasu Mitsuji

Olougouna Edgar

Ostendorf Mari

Park Kyu-Sik

Paulet Russell


Romano Giovanni

Seide Frank

Shah Dhavan V

Shinoda Koichi

Song Yeongkil

Stolcke Andreas

Suo Hongbin

Thilliez M

Wang Lan

Wang Wen

Westerhausen Rene

Woodland Philip C

Xie Xing

Yan Yonghong

Yap Kim-Hui

Yeom Heon Y

Yi Xun



0 50 100 150 200 250

Number of topics

000

002

004

006

008

(a)

0 50 100 150 200 250

Number of topics

1500

2000

2500

3000

Perp

lexi

ty

3500

(b)




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia




Topi

c_39

Topi

c_38

Topi

c_37

Topi

c_35

Topi

c_34

Topi

c_33

Topi

c_32

Topi

c_31

Topi

c_30

Topi

c_29

Topi

c_28

Topi

c_27

Topi

c_26

Topi

c_25

Topi

c_24

Topi

c_23

Topi

c_22

Topi

c_21

Topi

c_20

Topi

c_19

Topi

c_18

Topi

c_16

Topi

c_15

Topi

c_14

Topi

c_13

Topi

c_12

Topi

c_10

Topi

c_9

Topi

c_8

Topi

c_7

Topi

c_6

Topi

c_3

Topi

c_2

Topi

c_5

Topi

c_17

Topi

c_40

Topi

c_1

Topi

c_4

Topi

c_36

Topi

c_11


4 Discussions








2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

2000

2005

2015

2010

00020406 Topic 1

00020406 Topic 2

00

02

04 Topic 3

000408

Topic 4

00020406 Topic 5

00

04

08 Topic 6

00020406 Topic 7

00020406 Topic 8

00

02

04 Topic 9

00020406 Topic 10

00020406 Topic 11

00020406 Topic 12

00020406 Topic 13

00020406 Topic 14

00020406 Topic 15

00020406 Topic 16


00

02

Topic 22

000

010

020 Topic 23

00020406 Topic 24

00

02

04

04



00

02

04

000

010

020

00

04

08

00

04

08

00

04

08

00

02

04

00

02

04

00

04

08

00020406

000408

00020406

00020406

00

04

08

00020406

00020406

00

04

08

00

04

08

00020406

00

02

04

00

04

08









































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



































5 Conclusions



Disclosure




Acknowledgments


References
























































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia

















































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


a bibliometric review of natural language processing

Documents