the erc csa calls 2008 and 2009 eracep project · pdf file · 2017-03-10the erc csa...
TRANSCRIPT
Dr. Thomas Reiss
Etienne Vignola-Gagné
Piret Kukk
Fraunhofer ISI Karlsruhe
Prof. Dr. Wolfgang Glänzel
Dr. Bart Thijs
KU Leuven
The ERC CSA Calls 2008 and 2009
ERACEP Project
European Research Council
Introduction
Objectives:
• Identification of topically emerging research areas and
analysis to what extent the activities supported by the
ERC cover and contribute to these research areas.
Concept:
• Combination of two perspectives:
• From the ‘landscape of science’
• From ERC funding procedures
│ 2
European Research Council
Concept
European Research Council
Landscape of Science
40 dynamic fields
Selection of 20 most dynamic fields
for fine mapping
Pilot with 6 fields
composed of 46 clusters
In-depth analysis
of 10 emerging clusters
Expert
validation
Identification of emerging fields
European Research Council
The bibliometric tool – a short guide
│ 5
1. Selecting subjects for fine mapping
2. Building the databases for different time slides
3. Structural epistemological mapping of
literature in each subject and time period
5. Cognitive/bibliometric
description of emerging fields
6. Providing input for
external use
Expert
validation
4. Identification of emerging topics
Expert
validation
User
WoS
European Research Council
1. Step – the dynamic fields
• Thomson Reuters ‘Web of Science’ Subject Categories
with striking growth patterns have been selected as starting
points in the sciences, social sciences and arts and
humanities.
• Growth alone does not yet ensure the ‘emergence’
character of the field, but may act as a marker of
excitement for a topic within research communities.
• These subjects were considered to have the potential to
cover emerging topics.
• In total, 20 categories were chosen for the ERACEP
project.
│ 6
European Research Council
│ 7
List of dynamic fields
(ordered by sharp ratio)
SCI
Factor 03-08 Mean 03-08
Standard
Deviation
03-08
Sharp Index
03-08
Nursing 1,71 22,38 9,46 0,15
Orthopaedics 0,59 9,78 2,34 0,14
Environmental sciences 0,55 9,20 2,60 0,11
Operations research & management science 0,70 11,25 4,08 0,11
Obstetrics & gynaecology 0,40 7,02 1,40 0,10
Medical ethics 1,10 16,33 9,17 0,09
Robotics 0,57 9,44 3,44 0,09
Engineering, biomedical 0,60 9,84 3,84 0,09
Energy & fuels 0,62 10,18 4,14 0,09
Oncology 0,34 5,96 0,82 0,08
Biotechnology & applied microbiology 0,45 7,75 2,26 0,08
Material sciences, characterisation & testing 0,54 9,10 3,69 0,07
Behavioural sciences 0,59 9,76 5,07 0,06
SSCI
Public, environmental & occupational health 0,75 0,12 0,04 11,16
Transportation 1,20 0,17 0,10 8,16
Geography 0,81 0,13 0,06 7,88
Psychology, experimental 0,44 0,08 0,03 2,72
AHCI
Standard
Deviation
98-08
Archaeology 0,56 0,10 0,09 4,86
Religion 0,41 0,07 0,07 4,62
Architecture 0,38 0,07 0,06 4,30
1. Step – the dynamic fields
European Research Council
Interpretation of sharp ratio measurements:
• Comparatively, greater increase in publications
assigned to these fields over the period of time
examined.
• Two possible explanations:
1. Individual researchers are producing more papers per annum
than before, and than other fields;
2. There are relatively more individual researchers publishing in
this field than before and comparatively with other fields
• In any case: these fields are becoming more interesting
within the scientific community
• Validation through 21 semi-directed interviews
│ 8
1. Step – the dynamic fields
European Research Council
2. Step – the database
• All bibliographic data of relevant documents
(Articles, Proceeding papers, Reviews) were
downloaded and collected.
• Bibliographic data needed are
1. document identifier,
2. title, keywords, abstracts and
3. references.
• Title, keywords, abstracts are used for textual
description and text similarity.
• References are used for link analysis.
│ 9
European Research Council
3. Mapping of literature in each subject and time
period
• A textual similarity among documents is assumed if the
same terms and phrases are used.
• (Citation-)link similarity is assumed if papers are jointly
cited or the same references are used.
│ 10
Bibliographic coupling according
to Garfield (2001)
European Research Council
3. Mapping of literature in each subject and time
period
Techniques for detecting new emerging yet coherent
structures (Background)
Topics are searched in the mirror of their scholarly literature.
• Growing frequency of specific terms within a given research
area.
• Extracted terms can be used for labelling and describing the
obtained clusters.
LAMIREL ET AL., IASTED – AIA, 2008
• Topological measures to determine the role of each paper in
the citation network to decide whether there are emerging
clusters.
SHIBATA ET AL., Technovation, 2008
│ 11
European Research Council
3. Mapping of literature in each subject and time
period
• Two popular approaches (text- and citation-based methods)
are in a way complementing each other.
• Braam et al. and Zitt & Bassecoulard were among the firsts
advocating the combination of the two methods.
BRAAM ET AL., JASIS, 1991
ZITT & BASSECOULARD, Scientometrics, 1994
• The superiority of such hybrid methods over text-based and
link-based approaches have been shown.
GLENISSON ET AL., JASIST, 2005
BOYACK AND KLAVANS, IPM, 2010
LIU ET AL., JASIST, 2010
│ 12
European Research Council
3. Mapping of literature in each subject and time
period
Structural analysis of the discipline (‘growth field’)
• We proceed from two non-overlapping periods: 1999-2003 and
2004-2008.
• The emerging topics are sought within selected subject
categories according to Thomson Reuters’ Journal Citation
Reports (JCR) with remarkable growth (according to the ‘sharp
ratio’).
• A combined text- and link-based approach was used for
clustering.
• The textual component is based on term frequencies, where
terms were extracted from and stemmed for titles and abstracts.
Keyword phrases are kept and stop words are removed.
• The citation-based component uses bibliographic coupling.
│ 13
European Research Council
3. Mapping of literature in each subject and time
period
• Different areas in the sciences, social sciences and
humanities have different cultures of giving and using
citations, as well as of using natural language and specific
(technical) terms and phrases.
│ 14
Therefore, a combined
textual and citation-based
cluster analysis has been
applied to two different time
slides within the full ten-year
period 1999-2008.
European Research Council
4. Identification of emerging topics
• The content of clusters and their evolution has been
analysed.
• A cluster has been considered to represent an emerging
topic if the topic has reached a certain ‘critical mass’ and
one of the following criteria is met.
I. Existing cluster with an exceptional growth with
regard to the second period,
II. Completely new cluster with its root in other clusters
in the previous period and
III. Existing cluster with a topic shift in the new time
period
│ 15
European Research Council
4. Identification of emerging topics
• In local clustering (e.g., of smaller disciplines) the common
vocabulary often distorts the lexical approach.
• Furthermore, terms are not specific to the same extent in
all research topics, therefore clusters have be labelled on
the basis of the terms from the lexical component and what
we called ‘core documents’.
• Core documents are considered to represent sets of
related literature and clusters independently of any
particular lexical description.
• Core documents can be determined even without
clustering and can thus be applied to different cluster
solutions.
│ 16
European Research Council
4. Identification of emerging topics
Core documents are documents that have a given number n
of strong (hybrid text-citation) links with other documents in
the cluster. The strength is expressed by the similarity
measure also used for the clustering.
GLÄNZEL & CZERWON, Scientometrics, 1996
GLÄNZEL & THIJS, Scientometrics, 2011
• Core documents represent the most important nodes in the
underlying document networks (i.e., the clusters under
study).
• Core documents provide an ‘retrieval effect’ as added
value; following all links and paths outgoing from core
documents covers all related documents and thus a large
part of the structure under study.
│ 17
European Research Council
4. Identification of emerging topics
│ 18
Visualisation of the link
environment of a `core
document'
European Research Council
4. Identification of emerging topics
Examples for Types I (exceptional growth) and II (new)
│ 19
Type II
Type I
Type II
engineering, biomedical environmental sciences
European Research Council
We introduce the following four examples.
│ 20
WoS Subject Category Emerging topic
environmental sciences nano-pollution
public, environmental & occupational health environmental factors
energy & fuels biofuel
biomedical engineering brain-machine-interface
4. Identification of emerging topics
European Research Council
Example: Selected core documents in “nano pollution”
│ 21
On colloid retention in saturated porous media in the presence of energy barriers: The failure of alpha, and opportunities to predict eta
The significance of heterogeneity on mass flux from DNAPL source zones: An experimental investigation
C-60 colloid formation in aqueous systems: Effects of preparation method on size, structure, and surface, charge
Individual and mixture effects of selected pharmaceuticals and personal care products on the marine phytoplankton species Dunaliella tertiolecta
Nanomaterials as possible contaminants: the fullerene example
Exploring e-waste management systems in the United States
Influence of electrolyte species and concentration on the aggregation and transport of fullerene nanoparticles in quartz sands
Toxicity of aqueous fullerene in adult and larval Fundulus heteroclitus
Effects of particle composition and species on toxicity of metallic nanomaterials in aquatic organisms
Precise and Accurate Compound Specific Carbon and Nitrogen Isotope Analysis of Atrazine: Critical Role of Combustion Oven Conditions
5. Cognitive/bibliometric description of emerging
topics
European Research Council
Publication activity and international collaboration in environmental
sciences (left) and ‘nano pollution’ (right)
│ 22
BEL
BRA
DNK
FRA
DEU
GRC
IND
ITA
JPN
NLD
CHN
POL
ESP
SWE
GBR USA
BEL
BRA
DNK
FRA
DEU
GRC
IND
ITA
JPN
NLDCHN
POL
ESP
SWEGBR
USA
5. Cognitive/bibliometric description of emerging
topics
European Research Council
Selected core documents in “environmental factors” within “public health”
│ 23
Mortality in 13 French cities during the August 2003 heat wave
Ambient carbon monoxide may influence heart rate variability in subjects with coronary artery disease
Temperature and mortality among the elderly in the United States - A comparison of epidemiologic methods
Effects of air pollution on heart rate variability: The VA Normative Aging Study
Association of air pollution with increased incidence of ventricular tachyarrhythmias recorded by implanted cardioverter defibrillators
The estimation of SARS incubation distribution from serial interval data using a convolution likelihood
Mortality displacement of heat-related deaths - A comparison of Delhi, Sao Paulo, and London
Multipoint linkage analysis for a very dense set of markers
Association of ventricular arrhythmias detected by implantable cardioverter defibrillator and ambient air pollutants in the St Louis, Missouri metropolitan area
Focused exposures to airborne traffic particles and heart rate variability in the elderly
5. Cognitive/bibliometric description of emerging
topics
European Research Council
Publication activity and international collaboration in ‘public health’ (left)
and ‘environmental factors’ (right)
│ 24
BEL
BRA
DNK
FRA
DEU
GRC
IND
ITA
JPN
NLD
CHN
POL
ESP
SWE
GBR
USA
BEL
BRA
DNK
FRA
DEU
GRC
IND
ITA
JPN
NLD
CHN
POL
ESP
SWE
GBR
USA
5. Cognitive/bibliometric description of emerging
topics
European Research Council
5. Cognitive/bibliometric description of emerging
topics
Selected core documents in “biofuel” within “energy & fuels”
│ 25
Process optimization for biodiesel production from mahua (Madhuca indica) oil using response surface methodology
Continuous production of biodiesel via transesterification from vegetable oils in supercritical methanol
Temperature effects on biohydrogen production in a granular sludge bed induced by activated carbon carriers
Biohydrogen generation from jackfruit peel using anaerobic contact filter
Biohydrogen-production from beer lees biomass by cow dung compost
Isolation of hydrogen generating microflora from cow dung for seeding anaerobic digester
Fermentative hydrogen production from xylose using anaerobic mixed microflora
Sulfate effect on fermentative hydrogen production using anaerobic mixed microflora
Biodiesel production via non-catalytic SCF method and biodiesel fuel characteristics
Biological hydrogen production in suspended and attached growth anaerobic reactor systems
European Research Council
Publication activity and international collaboration in ‘energy & fuels’ (left)
and ‘biofuel’ (right)
│ 26
BEL
BRA
DNK
FRA
DEU
GRC
IND
ITA
JPN
NLD
CHN
POL
ESP
SWE
GBR
USA
BEL
BRA
DNK
FRA
DEU
GRC
IND
ITA
JPN
NLD
CHN
POL
ESPSWE
GBR
USA
5. Cognitive/bibliometric description of emerging
topics
European Research Council
Example: Selected core documents in “brain-machine interface”
│ 27
"Virtual keyboard" controlled by spontaneous EEG activity
Planar gradiometer for magnetic induction tomography (MIT): theoretical and experimental sensitivity maps for a low-contrast phantom
Adaptive BCI based on variational Bayesian Kalman filtering: An empirical evaluation
Model-based neural decoding of reaching movements: A maximum likelihood approach
Ascertaining the importance of neurons to develop better brain-machine interfaces
Anasynchronously controlled EEG-based virtual keyboard: Improvement of the spelling rate
Boosting bit rates in noninvasive EEG single-trial classifications by feature combination and multiclass paradigms
Support vector channel selection in BCI
Classification of single-trial electroencephalogram during finger movement
BCI2000: A general-purpose, brain-computer interface (BCI) system
5. Cognitive/bibliometric description of emerging
topics
European Research Council
Publication activity and international collaboration in biomedical
engineering (left) and ‘brain-machine interface’ (right)
│ 28
BEL
BRA
DNK
FRA
DEU
GRC
IND
ITA
JPN
NLDCHN
POL
ESP
SWE
GBR
USA
BEL
BRA
DNK
FRA
DEU
GRC
IND
ITA JPNNLD
CHN
POL
ESP
SWE
GBR
USA
5. Cognitive/bibliometric description of emerging
topics
European Research Council
Observations
Publication activity and international collaboration
reflect globalisation and regional/national aspects.
• Measurable differences in ‘environmental factors’, ‘nano
pollution’, ‘bio diesel’ and ‘brain-machine interface’ found
• Essential contribution of the EU in these topics
• Growing role of the emerging economies in Asia and
South America
│ 29
5. Cognitive/bibliometric description of emerging
topics
European Research Council
│ 30
5. Cognitive/bibliometric description of emerging
topics
Cluster evolution of the discipline energy & fuels
energy & fuels
Legend:
red: 1999-2003
green: 2003-2008
blue: 2009-2011
European Research Council
7. Conclusions
│ 31
1. Selecting subjects for fine mapping
• The proposed methodolgy using sharp ratio’s is working. However,
the method is limited to the subject category assignment done by
the data provider.
• The selection of “growth fields” is not mandatory. The dynamics of
these fields just facilitated the development of methodology, which,
however, works with all fields.
3. Structural epistemological mapping of
literature in each subject and time period
• The hybrid approach proved to be very successfull in clustering
the given fields.
European Research Council
7. Conclusions
│ 32
4. Identification of emerging topics
• The application of two independent clustering and the citation links
between two periods revealed the dynamic structures in the fields.
• Three types of evolution are relevant to this project.
• Emerging topics could be identified in several subject categories.
• Expert validation confirmed our findings.
• The experimental extension of the time window to 2009-2011 has
confirmed the stability of the approach.
European Research Council
7. Conclusions
│ 33
• Emerging fields are described by keywords and by core documents.
These core documents give the best description.
• Other bibliometric indicators can be used, e.g., national publication
activity and international collaboration.
5. Cognitive/bibliometric
description of emerging fields
European Research Council
Appendix
│ 34
Publication of ERACEP results in international scientific journals
Glänzel, W., Thijs, B. (2011), Using `core documents' for the representation of clusters and
topics. Scientometrics, 88(1), 297-309.
Glänzel, W., Thijs, B. (2011), Research in emerging fields: who takes the lead? ISSI
Newsletter, 7(4), 85-95.
Meyer, M., Libaers, D. Thijs, B., Glänzel, W. (2011), The emergence of entrepreneurship as
a research field. In: E. Noyons et al. (Eds), Proceedings of ISSI 2011, Durban, South
Africa, 552-557.
Glänzel, W. (2012), Bibliometric methods for detecting and analysing emerging research
topics. El Profesional de la Información, 21(2), 194-201.
Glänzel, W., Thijs, B. (2012), Using ‘core documents’ for detecting and labelling new
emerging topics. Scientometrics, 91(2), 399-416.
Glänzel, W. (2012), The role of core documents in bibliometric network analysis and their
relation with h-type indices. Scientometrics, 93(1), 113-123.
Meyer, M., Thijs, B., Libaers, D., Glänzel, W., Granta, K., Debackere, K. (2012), On the
Origin and Emergence of Entrepreneurship as a Research Field. In: E. Archambault,
Y. Gingras, V. Lariviere (eds), Proceedings of STI 2012 Montreal, Vol. 2, 879-880.
Thijs, B., Schiebel, E. Glänzel, W. (2013), Do second-order similarities provide added-value
in a hybrid approach? Scientometrics, DOI: 10.1007/s11192-012-0896-1, in press.