Second Workshop on Theory-Informed User Modeling for Tailoring and Personalizing Interfaces (HUMANIZE): Workshop Preface Mark Graus Bruce Ferwerda Marko Tkalcic Human-Technology Department of Computer Faculty of Computer Science Interaction Group Science and Informatics Free University of Eindhoven University of School of Engineering Bozen-Bolzano Technology Jönköping University Piazza Domenicani 3 P.O. Box 513 P.O. Box 1026 I-39100, Bozen-Bolzano, IT 5600 MB, Eindhoven, NL SE-551 11, Jönköping, SE marko.tkalcic@unibz.it m.p.graus@tue.nl bruce.ferwerda@ju.se Panagiotis Germanakos∗ UX, Mobile & Business Services, P&I ICD SAP SE Dietmar-Hopp-Allee 16 69190, Walldorf, DE panagiotis.germanakos@sap.com ABSTRACT their needs in order to provide the optimal interface for its The second workshop on Theory-Informed User Modeling for users. When creating user interfaces that can be personalized, Tailoring and Personalizing Interfaces (HUMANIZE)1 took quite often a more data-driven approach is taken, where practi- place in conjunction with the 23rd annual meeting of the in- tioners rely on methods that use implicit or explicit feedback telligent user interfaces (IUI)2 community in Tokyo, Japan to prescribe how to alter an interface. on March 11, 2018. The goal of the workshop was to attract The current workshop aims at soliciting work that investigates researchers from different fields by accepting contributions the potential of combining the more practical data mining/ma- on the intersection of practical data mining methods and theo- chine learning methods with a more theory-driven approach. retical knowledge for personalization. A total of eight papers Three aspects play an important role in taking a more theory- were accepted for this edition of the workshop. driven approach to personalization: Author Keywords 1. How to consider the users of a system and their individual User modeling, personalization, tailoring, user interfaces differences. 2. How to infer these individual differences from interaction INTRODUCTION data. When designing interfaces practitioners often rely on knowl- 3. How to translate individual differences into interface de- edge and experience about the interface’s intended users and signs. * Also affiliated with the Department of Computer Science, Uni- versity of Cyprus, P.O. Box 20537, 1678, Nicosia, Cyprus, pger- The characteristics that play a role in what a user needs or man@cs.ucy.ac.cy wants from a system need to be investigated. Knowing what 1 https://humanize2018.wordpress.com/ users differ on allows us to alter the interface. These charac- 2 http://iui.acm.org/2018/ teristics can then be used to construct a user model containing this information. Examples of characteristics that may play a role in how to design an optimal interface are cognitive style, personality, and susceptibility to persuasive strategies. Secondly, there is a challenge of profiling a user in terms of these characteristic based on interaction data. Several ap- proaches exist for this more computational challenge, for ex- ©2018. Copyright for the individual papers remains with the authors. ample mining data from social media and clickstream analysis. Copying permitted for private and academic purposes. HUMANIZE ’18, March 11, 2018, Tokyo, Japan A third aspect is knowing how to adapt an interface to match a REFERENCES certain type of user. When a user’s characteristics are known, 1. Bruce Ferwerda and Marko Tkalcic. 2018. You Are What the interface can be altered to match this user. For example You Post: What the Content of Instagram Pictures Tells by reducing the number of search results for users under high About Users’ Personality. In Companion Proceedings of cognitive load, or manipulating diversity. the 23rd International on Intelligent User Interfaces: 2nd Workshop on Theory-Informed User Modeling for These challenges are interconnected and there is no natural Tailoring and Personalizing Interfaces (HUMANIZE). order in which these aspects need to be addressed when per- sonalizing an interface. For example, by analyzing behavior 2. Mark Graus, Martijn Willemsen, and Chris Snijders. data we can identify potential individual characteristics that 2018. Personalizing a parenting app: parenting-style play a role in people’s needs. surveys beat behavioral reading-based models. In Companion Proceedings of the 23rd International on HUMANIZE provides scholars and practitioners in the field Intelligent User Interfaces: 2nd Workshop on of personalized user interfaces with a venue to discuss and Theory-Informed User Modeling for Tailoring and explore the commonalities between the sub-problems involved Personalizing Interfaces (HUMANIZE). with user interface personalization. 3. Maciej J Korzepa, Benjamin Johansen, Michael K An non-exhaustive list of topics for this workshop: Petersen, Jan Larsen, Jakob E Larsen, and Niels H • Identifying models that are (expected to be) useful for per- Pontoppidan. 2018. Learning preferences and sonalizing user interfaces (e.g., personality, level of domain soundscapes for augmented hearing. In Companion knowledge, need for cognition, cognitive styles) Proceedings of the 23rd International on Intelligent User • Data mining methods to infer user profiles in terms of cogni- Interfaces: 2nd Workshop on Theory-Informed User tive/psychological user characteristics from data (e.g., how Modeling for Tailoring and Personalizing Interfaces to infer personality from social media or domain knowledge (HUMANIZE). from clickstreams) 4. Johannes Kunkel, Tim Donkers, Catalin-Mihai Barbu, • Theory on how to tailor interfaces to better match certain and Jurgen Ziegler. 2018. Trust-related Effects of user profiles (e.g., altering the number of search results, Expertise and Similarity Cues in Human-Generated ordering of interface elements, visual versus textual repre- Recommendations. In Companion Proceedings of the sentations) 23rd International on Intelligent User Interfaces: 2nd • User studies investigating one or more of the aspects men- Workshop on Theory-Informed User Modeling for tioned above Tailoring and Personalizing Interfaces (HUMANIZE). CONTRIBUTIONS 5. Alixe Lay and Bruce Ferwerda. 2018. Predicting users’ A total of eight papers were accepted: 3 long papers, 4 short personality based on their ’liked’ images on Instagram. In papers, and 1 position paper. Papers were categorized into one Companion Proceedings of the 23rd International on of the three sessions: 1) Personality, 2) Social, or 3) Health & Intelligent User Interfaces: 2nd Workshop on Wellbeing. Below a description of the accepted papers: Theory-Informed User Modeling for Tailoring and Personalizing Interfaces (HUMANIZE). Personality. Lay and Ferwerda [5] proposes a new view on how to incorporate meta data of Instagram users to infer 6. Yu Xu and Michael J Lee. 2018. Shopping as a Social their personality traits. Similarly, Ferwerda and Tkalcic [1] Activity: Understanding People’s Categorical Item analyzed the content of Instagram pictures and found distinct Sharing Preferences on Social Networks. In Companion correlations with personality traits. Zheng [7] on the other Proceedings of the 23rd International on Intelligent User hand investigated how personality traits influence individual Interfaces: 2nd Workshop on Theory-Informed User and group decision making. Modeling for Tailoring and Personalizing Interfaces (HUMANIZE). Social. Xu and Lee [6] explored what kind of products people choose to share on their social networks that they have bought 7. Yong Zheng. 2018. Identifying Dominators and online. Kunkel et al. [4] compared the effect of personal Followers in Group Decision Making Based on The and impersonal recommendation sources, and investigated the Personality Traits. In Companion Proceedings of the 23rd influence of traits of personal recommendation sources on a International on Intelligent User Interfaces: 2nd user’s trust in recommendations Workshop on Theory-Informed User Modeling for Tailoring and Personalizing Interfaces (HUMANIZE). Health & Wellbeing Korzepa et al. [3] describes how to use behavioral data for personalized hearing aid profiles. Zhou et 8. Mo Zhou, Yonatan Mintz, Yoshimi Fukuoka, Ken al. [8] are using reinforcement learning to generate personal- Goldberg, Elena Flowers, Philip Kaminsky, Alejandro ized motivators for fitness applications that are challenging but Castillejo, and Anil Aswani. 2018. Personalizing Mobile attainable. Graus et al. [2] showed that personalization based Fitness Apps using Reinforcement Learning. In on parenting styles gained a higher perceived personalization Companion Proceedings of the 23rd International on and satisfaction than reading-based personalization. Intelligent User Interfaces: 2nd Workshop on Theory-Informed User Modeling for Tailoring and Personalizing Interfaces (HUMANIZE).