ubittention 2020: 5th international workshop on …...research, context-aware and ubiquitous...
TRANSCRIPT
UbiTtention 2020: 5th International Workshop onSmart & Ambient Notification and Attention
ManagementNiels van BerkelAalborg UniversityAalborg, Denmark
Anja ExlerKarlsruhe Institute of Technology
Karlsruhe, [email protected]
Martin GjoreskiJozef Stefan InstituteLjubljana, Slovenia
Tine KolenikJozef Stefan InstituteLjubljana, [email protected]
Tadashi OkoshiKeio UniversityTokyo, Japan
Veljko PejovicUniversity of LjubljanaLjubljana, Slovenia
Aku VisuriUniversity of Oulu
Oulu, [email protected]
Alexandra Voitadesso SE
Dortmund, [email protected]
ABSTRACTUsers are increasingly confronted with a tremendous amountof information proactively andwithout explicit consent throughnotifications from a variety of applications and services. Thisinformation load is increased due to the ubiquity of end-user(mobile) devices. Novel computing paradigms such as IoT andsmart cities may further overload end-users, despite the clearindication from literature that human attention is limited. Tocounter this challenge, “attention management”, includingattention representation, sensing, prediction, analysis, per-sonalization, and adaptive behavior is needed in our comput-ing systems. Following the successful UbiTtention workshopseries as organised from 2016 on-wards, the UbiTtention2020 workshop brings together researchers and practitionersfrom academia and industry to explore the management ofhuman attention and notifications across versatile devicesand contexts. UbiTtention presents and elicits research toovercome information overload and overchoice – tailoringdevice or application behavior to user needs.
Permission to make digital or hard copies of all or part of this work forpersonal or classroom use is granted without fee provided that copies are notmade or distributed for profit or commercial advantage and that copies bearthis notice and the full citation on the first page. Copyrights for componentsof this work owned by others than ACMmust be honored. Abstracting withcredit is permitted. To copy otherwise, or republish, to post on servers or toredistribute to lists, requires prior specific permission and/or a fee. Requestpermissions from [email protected]/ISWC ’20 Adjunct, September 12–16, 2020, Cancun, Mexico© 2020 Association for Computing Machinery.ACM ISBN 978-1-4503-6869-8/19/09. . . $15.00https://doi.org/10.1145/3341162.3347766
CCS CONCEPTS•Human computer interaction (HCI); •Ubiquitous com-puting;
KEYWORDSNotifications, Attention Management, Ambient Interfaces,Smart Cities, Internet of Things, Interruption Management
ACM Reference Format:Niels van Berkel, Anja Exler, Martin Gjoreski, Tine Kolenik, TadashiOkoshi, Veljko Pejovic, Aku Visuri, and Alexandra Voit. 2020. UbiT-tention 2020: 5th International Workshop on Smart & AmbientNotification and Attention Management. In Adjunct Proceedingsof the 2020 ACM International Joint Conference on Pervasive andUbiquitous Computing and the 2020 International Symposium onWearable Computers (UbiComp/ISWC ’20 Adjunct), September 12–16,2020, Cancun, Mexico. ACM, New York, NY, USA, 5 pages. https://doi.org/10.1145/3341162.3347766
1 BACKGROUNDAttention is one of the most precious resources we have,yet, today’s ubiquitous computing environments indiscrimi-nately exploit this resource for individual applications’ ben-efits. Devices and services are proactively seeking our atten-tion through notifications: visual, auditory, or tactile alerts in-tended to draw attention to events that took place outside of auser’s focus [8]. Attention grabbing is rapidly becoming inte-gral to many ubiquitous computing platforms, such as smart-phones, desktop computers, and vehicles, where notificationsare used by multiple applications and services [18, 22], suchas social networks, email clients, or driving alert systems.
UbiComp/ISWC ’20 Adjunct, September 12–16, 2020, Cancun, Mexico van Berkel et al.
Despite the benefit of triggering a timely reaction, andindeed, potentially legitimate steering of a user’s attentionto critical information, research has repeatedly shown thatnotifications can be distracting, causing a negative effect ontask performance [3, 5, 7, 8, 12, 19], increasing stress [11],and reducing wellbeing [10]. Furthermore, often used bycommunication platforms [4, 6, 16, 18], notifications carrycertain social expectations towards responsiveness, result-ing in added pressure to attend notifications. Consequently,disabling notifications, despite their known disadvantages,for many people is not an option [6, 8, 12].At the core of the issue with distraction in ubiquitous
computing environments lies our limited cognitive capac-ity. Different cognitive loads demanded by the offline andonline tasks we perform, tasks signaled by notifications, aswell as tasks performed while the notifications are buzzing,make it challenging for the user to triage their attention effi-ciently. Eventually, this leads to the problems of informationoverload and overchoice—in our opinion two of the most rel-evant problems in information technology for the next fewdecades. In the era of the Internet of Things (IoT) we haveto handle incoming notifications from both our personal aswell as ‘other’ devices. Together with developments in smartcity environments or with smart mobility, this informationoverload is expected to grow.Recently, we have seen an increase in novel attempts to
address these problems by utilising contextual data [7, 9, 13,15, 17, 20] or notification content [14, 21] in order to delivernotifications at more opportune moments, the use of ambientinformation presentation or augmentation, or by making iteasier to deal with interruptions [2]. However, significant
Figure 1: This alarm clock uses ambient light to softly wakeup its user. We envision that it could also be used as anhelpful output device for other information, e.g., incomingphone calls or reminders.
Table 1: Overview of workshop topics.
(1) Detection/prediction of users’ status around attentionand notifications, such as availability, interruptibility,attentional status and cognitive load
(2) Exchange/share data, analysis and feedback on capa-bilities beyond detection and prediction
(3) Versatile types of information presentation methodsincluding ambient, peripheral, distributed and multi-modal presentation
(4) Infrastructures, frameworks and tools for the develop-ment of smart attention systems
(5) Strategies for attention management against emergingcomputing technologies, including IoT devices
(6) Understanding users’ behavior and habits around noti-fications and interruption, including longer term “userengagement” and “behavior change”
(7) Use of ambient representations for “big-data analysis”(8) Management of information overload in various
emerging computing venues such as smart city andsmart mobility
challenges, especially along the lines of understanding cog-nitive load and our information processing limitations [1], aswell as devising technologies that interact with their usersin adaptive multimodal ways, still remain.
2 UBITTENTION 2020 WORKSHOPIn this workshop, we will bring together people from indus-try and academia who are active in the areas of attentionresearch, context-aware and ubiquitous computing, person-alization and adaptation, and ambient and multi-modal in-teraction. The main objective of UbiTtention 2020 is to sharethe latest research on user attention and notification manage-ment in several research areas and contexts, including HCI,systems research, user studies, IoT, smart cities, smart homes,and smart vehicles (as shown in Table 1). Further, this work-shop aims to identify future research challenges, researchopportunities, and applications of our research outcomes tosociety at large.This workshop is a follow-on from the successful Smart-
tention, Please! workshops at MobileHCI 2015 and 2016, andthe UbiTtentionworkshops at UbiComp 2016, 2017, 2018, and20191. Last year, the UbiTtention workshop received 12 inno-vative papers and more than 35 participants on the venue ofUbiComp 2018. While the Smarttention workshops focusedmore on mobile notifications, the subsequent UbiTtentionworkshops focus on a larger understanding of the differentroles notifications can play in a wide variety of computing
1https://www.ubittention.org/2019/
UbiTtention 2020 UbiComp/ISWC ’20 Adjunct, September 12–16, 2020, Cancun, Mexico
Figure 2: Discussion at the UbiTention ’17 workshop at Ubi-Comp ’17 in Mauii.
Figure 3: Audience at the UbiTention ’18 workshop at Ubi-Comp ’18 in Singapore.
environments including the office, the home, in cars, as wellas other smart environments.
Relevance and impact to UbiComp: UbiTtention’s fo-cus on user attention and notifications is related to key re-search fields of UbiComp, such as HCI, mobile computing,IoT-based services, and others, thus the workshop has sig-nificant potential to impact and contribute to all the listedareas.
In addition, past UbiComp papers represent a cornerstoneof attention management research [1, 13, 15, 22], making Ubi-Comp 2020 the ideal venue for UbiTtention 2020 workshop.The impact is bidirectional, as the tradition of UbiTtentionbeing collocated with UbiComp has resulted in a number ofcontributions initially presented and shaped by discussionsat UbiTtention, ending up published in ACM IMWUT.
Finally, UbiComp is rapidly engaging machine learning re-searchers and practitioners interested in making real-world
impact. In line with this trend, UbiTtention 2020 introducesthe cognitive load inference challenge (see Section 5), thusproviding a collaboration platform for machine learning,ubiquitous computing, and psychology researchers. Our pre-liminary contacts with diverse research groups from academiaand industry, e.g. Dr. Garbas (Fraunhofer IIS), Dr. Lustrek (In-stitute Jozef Stefan – winners of last year’s UbiComp HASCAand Emteq challenges), and Dr David (Kassel University),who have expressed interest in participating in such a chal-lenge, point out that UbiTtention may promote UbiCompparticipation within a rather diverse audience.
Long-termobjectives:Theworkshopwill lead to a deeperunderstanding of cognitive load, notifications, and attentionmanagement. We plan to consolidate and significantly ex-pand an already vibrant research community formed aroundthe previous installments of UbiTtention. More specifically,we are already communicating with Springer publishing houseabout a book on the topic of attention management in ubiq-uitous computing, and UbiTtention will help us to identifycutting edge research contributions whose authors will beinvited to contribute chapters to the mentioned book.
3 WORKSHOP STRUCTUREWe plan UbiTtention 2020 as a full-day workshop, aiming tobring together practitioners and academics. The key partsof the workshop include: a keynote, a paper presentationsession, and an open-data challenge.
Previous years have demonstrated the benefits of keynotetalks that enable UbiTtention audience to gain deeper in-sights on a particular research achievement in the area ofattention management. Moreover, we found that the work-shop keynote format enables a larger degree of interactionbetween the presenter and the audience. We will strive tohave a female and/or a minority keynote. Potential candi-dates we plan to invite include Shamsi Iqbal, Gloria Mark,and Mary Czerwinski.The second part of the workshop consists of short pa-
per presentations. UbiTtention does not aim to be a mini-conference, thus, the paper presentations are not a goal perse, but a means to spark discussion with the audience. Thediscussions serve as an early feedback and brainstormingsession aimed to help presenters steer their future researchendeavors. UbiTtention papers will not be selected merely ontheir technical quality, but the likeliness of the topic sparkingdiscussion at the workshop. Papers describing an early-stagedemo products will be especially welcome.
Finally, a novel part of this year’s UbiTtention is the open-data challenge. The challenge, motivated by a discussionheld at UbiTtention 2019, where the lack of open datasetsused for the evaluation of different approaches towards at-tention management was highlighted, enables researchersto develop and evaluate innovative solutions in the area of
UbiComp/ISWC ’20 Adjunct, September 12–16, 2020, Cancun, Mexico van Berkel et al.
Table 2: The workshop will be a mixture of recent researchresult presentations, keynote talks, group discussions, and apresentation ofwinning entries for the open-data challenge.
09:00–09:10 Introductions09:10–10:00 Keynote Lecture10:00–10:30 Coffee Break10:30–12:30 Paper Presentation Session12:30–13:30 Lunch Break13:30–15:00 Open-Data Challenge Presentations15:00–15:30 Coffee Break15:30–16:00 Challenge Winner Announcement16:00–17:00 Wrap-up Session & Plan of Future Actions18:30–20:00 Workshop Dinner
cognitive load inference, one of the key sub-areas supportingattention management in ubiquitous computing.
4 WORKSHOP PAPERSOur workshop will ask for papers outlining new insights inthe topics listed in Table 1. The papers are written in theSIGCHI Extended Abstract format with a length of 4 to 6pages and are to be included in the ACM Digital Library.Call for papers will be sent to all relevant mailing lists ofseveral research communities, including those of UbiCompresearch, HCI, mobile interaction, as well as the IoT, smartcity, smart home, and smart vehicle communities. Further, thecall will be distributed to participants of the previous Smart-tention and UbiTtention workshops. Finally, information willalso be shared on the workshop website and Facebook page.We aim for a selection of 10-15 workshop papers. Submis-sions will be peer-reviewed by at least two reviewers froman international PC of renowned researchers and experts.The acceptance criteria will be a mixture of relevance, nov-elty, writing, research quality, and provocativeness. With theaforementioned measures, we are confident to attract a largenumber of submissions, which will allow us to assemble ahigh-quality workshop.
5 OPEN-DATA CHALLENGETo prevent undesirable effects of attention grabbing at timeswhen a user is occupied with a difficult task, ubiquitous com-puting devices should be aware of the user’s cognitive load.When humans experience a psycho-physiological load, e.g.,in the form of a demanding task, the activation of the sym-pathetic nervous system increases. The increased activationsubstantially speeds up certain processes within the body(‘fight-or-flight’ response). It raises the heart rate, sweatingrate, breathing rate, and blood pressure; the pupils dilate;the saliva flow decreases; the heart-beats become equidis-tant; the blood flow is restricted from the extremities, andis redirected towards the vital organs. Traditionally, these
signals have been measured in highly controlled environ-ments using specialized equipment. Recently, the desire forin-situ knowledge of cognitive load has led to advances inphysiological signal-based cognitive load inference usingcommodity devices. However, inferring cognitive load is ex-tremely challenging, especially when performed withoutobtrusive, expensive, and purpose-built equipment.
To advance the field of cognitive load inference in ubiqui-tous computing, we will organize a public machine learning(ML) challenge in which competitors can build their own MLpipelines for cognitive load inference. All ML pipelines willbe trained and evaluated on the same dataset, thus enablinga systematic and comparable advancement. The dataset isthe first labelled dataset for cognitive load monitoring witha wristband and it will be fully released after the challenge.During the challenge, the participants will only have accessto a labelled training subset and unlabelled test subset. Thelabels of the test subset will be kept in secret. The dataset wascollected in a quiet, normal-temperature roomwhile workingon a PC with the wristband strapped to her non-dominantarm. Twenty participants attended these experiments. Thetechnical details of the challenge are described in the call forparticipation.
6 CONCLUSIONThe successful UbiTtention series of workshops on user at-tention and notification management has provided valuablelessons both for academia and industry in the fields such asHCI, mobile computing, and IoT-based services. The dynamicdevelopment of these fields seeks for continuous and activeaddressing of the related research challenges and opportuni-ties. We expect that approximately 35 participants, leadingUbiComp machine learning teams, and organizers who haveled the community for the past five years, will ensure thatthis year’s UbiTtention showcases the latest research on userattention and notification management, thus empoweringfurther growth and enabling transition of this knowledgeinto applications that can positively impact the society.
REFERENCES[1] Christoph Anderson, Isabel Hübener, Ann-Kathrin Seipp, Sandra Ohly,
Klaus David, and Veljko Pejovic. 2018. A survey of attention manage-ment systems in ubiquitous computing environments. Proceedings ofthe ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies2, 2 (2018), 1–27.
[2] Matthias Böhmer, Christian Lander, Sven Gehring, Duncan P. Brumby,and Antonio Krüger. 2014. Interrupted by a Phone Call: ExploringDesigns for Lowering the Impact of Call Notifications for SmartphoneUsers. In CHI ’14 (Toronto, Ontario, Canada). ACM.
[3] Jelmer P. Borst, Niels A. Taatgen, and Hedderik van Rijn. 2015. WhatMakes Interruptions Disruptive?: A Process-Model Account of theEffects of the Problem State Bottleneck on Task Interruption and Re-sumption. In CHI ’15 (Seoul, Republic of Korea). ACM.
UbiTtention 2020 UbiComp/ISWC ’20 Adjunct, September 12–16, 2020, Cancun, Mexico
[4] Karen Church and Rodrigo de Oliveira. 2013. What’s Up with What-sapp?: Comparing Mobile Instant Messaging Behaviors with Tradi-tional SMS. In MobileHCI ’13 (Munich, Germany). ACM.
[5] Mary Czerwinski, Eric Horvitz, and Susan Wilhite. 2004. A DiaryStudy of Task Switching and Interruptions. In CHI ’04 (Vienna, Austria).ACM.
[6] Nitesh Goyal and Susan R. Fussell. 2016. Effects of SensemakingTranslucence on Distributed Collaborative Analysis. In Proceedings ofthe 19th ACM Conference on Computer-Supported Cooperative Work &Social Computing (San Francisco, California, USA) (CSCW ’16). ACM,New York, NY, USA, 288–302.
[7] Nitesh Goyal and Susan R. Fussell. 2017. Intelligent Interruption Man-agement Using Electro Dermal Activity Based Physiological Sensorfor Collaborative Sensemaking. Proc. ACM Interact. Mob. WearableUbiquitous Technol. 1, 3, Article 52 (Sept. 2017), 21 pages.
[8] Shamsi T. Iqbal and Eric Horvitz. 2010. Notifications and Awareness:A Field Study of Alert Usage and Preferences. In CSCW ’10 (Savannah,Georgia, USA). ACM.
[9] Yasumasa Kobayashi, Takahiro Tanaka, Kazuaki Aoki, and Kinya Fujita.2015. Automatic Delivery Timing Control of Incoming Email Based onUser Interruptibility. In CHI EA ’15 (Seoul, Republic of Korea). ACM.
[10] Kostadin Kushlev, Jason Proulx, and Elizabeth W Dunn. 2016. " SilenceYour Phones" Smartphone Notifications Increase Inattention and Hy-peractivity Symptoms. In Proceedings of the 2016 CHI conference onhuman factors in computing systems. 1011–1020.
[11] Gloria Mark, Daniela Gudith, and Ulrich Klocke. 2008. The cost ofinterrupted work: more speed and stress. In Proceedings of the SIGCHIconference on Human Factors in Computing Systems. 107–110.
[12] Gloria Mark, Stephen Voida, and Armand Cardello. 2012. "A Pace NotDictated by Electrons": An Empirical Study of Work Without Email.In CHI ’12 (Austin, Texas, USA). ACM.
[13] Tadashi Okoshi, Julian Ramos, Hiroki Nozaki, Jin Nakazawa, Anind K.Dey, and Hideyuki Tokuda. 2015. Reducing Users’ Perceived Men-tal Effort Due to Interruptive Notifications in Multi-device Mobile
Environments. In UbiComp ’15 (Osaka, Japan). ACM.[14] Tadashi Okoshi, Kota Tsubouchi, Masaya Taji, Takanori Ichikawa, and
Hideyuki Tokuda. 2017. Attention and Engagement-Awareness in theWild : A Large-Scale Study with Adaptive Notifications. In PerCom ’17(Maui, Hawaii). IEEE.
[15] Veljko Pejovic and Mirco Musolesi. 2014. InterruptMe: Designing Intel-ligent Prompting Mechanisms for Pervasive Applications. In UbiComp’14 (Seattle, Washington). ACM.
[16] Martin Pielot, Karen Church, and Rodrigo de Oliveira. 2014. An In-situStudy of Mobile Phone Notifications. In MobileHCI ’14 (Toronto, ON,Canada). ACM.
[17] Benjamin Poppinga, Wilko Heuten, and Susanne Boll. 2014. Sensor-Based Identification of Opportune Moments for Triggering Notifica-tions. IEEE Pervasive Computing (2014).
[18] Alireza Sahami Shirazi, Niels Henze, Tilman Dingler, Martin Pielot,Dominik Weber, and Albrecht Schmidt. 2014. Large-scale Assessmentof Mobile Notifications. In CHI ’14 (Toronto, Ontario, Canada). ACM.
[19] Cary Stothart, Ainsley Mitchum, and Courtney Yehnert. 2015. Theattentional cost of receiving a cell phone notification. Journal ofexperimental psychology: human perception and performance (2015).
[20] Niels van Berkel, Jorge Goncalves, Peter Koval, Simo Hosio, TilmanDingler, Denzil Ferreira, and Vassilis Kostakos. 2019. Context-InformedScheduling and Analysis: Improving Accuracy of Mobile Self-Reports.In Proceedings of ACM SIGCHI Conference on Human Factors in Com-puting Systems (CHI’19). 51:1–51:12.
[21] Aku Visuri, Niels van Berkel, Tadashi Okoshi, Jorge Goncalves, andVassilis Kostakos. 2019. Understanding Smartphone Notifications’User Interactions and Content Importance. International Journal ofHuman-Computer Studies 128 (2019), 72–85.
[22] DominikWeber, Alexandra Voit, Philipp Kratzer, and Niels Henze. 2016.In-situ Investigation of Notifications in Multi-device Environments. InUbicomp ’16 (Heidelberg, Germany). ACM.