A Summary of the Third Workshop on Theory-Informed User Modeling for Tailoring and Personalizing Interfaces Mark P. Graus∗ Bruce Ferwerda School of Business and Economics Department of Computer Science and Informatics Maastricht University Jönköping University Maastricht, the Netherlands Jönköping, Sweden mp.graus@maastrichtuniversity.nl bruce.ferwerda@ju.se Marko Tkalčič Panagiotis Germanakos† Faculty of Computer Science UX ICD, Product Engineering, Intelligent Enterprise Group Free University of Bozen-Bolzano SAP SE Bozen-Bolzano, Italy Walldorf, Germany marko.tkalcic@unibz.it panagiotis.germanakos@sap.com ABSTRACT The current workshop aims at soliciting work that investigates The third workshop on Theory-Informed User Modeling for Tailoring the potential of combining the more data-driven data mining/machine and Personalizing Interfaces (HUMANIZE)1 took place in conjunc- learning methods with a more theory-driven user-centered ap- tion with the 24th annual meeting of the intelligent user interfaces proach. Incorporating more knowledge and theory when applying (IUI)2 community in Los Angeles, CA, USA on March 20, 2019. The personalization in systems can lead to increased performance of goal of the workshop was to attract researchers from different fields these systems. For example, taking a user’s cognitive style into by accepting contributions on the intersection of practical data min- account when personalizing an online tool for comparing and pur- ing methods and theoretical knowledge for personalization. A total chasing mobile phone contracts, can increase the purchase intent of six papers were accepted for this edition of the workshop. of these systems [2]. Three main aspects play an important role in taking a more CCS CONCEPTS theory-driven approach to personalization: • Information systems → Data mining; • Human-centered (1) How to consider the users of a system in terms of user charac- computing → HCI theory, concepts and models; Interaction design teristics and individual differences on these characteristics? process and methods. (2) How to infer these user characteristics and individual differ- ences from interaction data? KEYWORDS (3) How to use these individual differences to personalize interfaces User modeling; personalization; tailoring; user interfaces or other aspects of systems? ACM Reference Format: A first challenge for addressing those research dimensions re- Mark P. Graus, Bruce Ferwerda, Marko Tkalčič, and Panagiotis Germanakos. lates to the characteristics that play a role in what users need or 2019. A Summary of the Third Workshop on Theory-Informed User Model- want from a system. Knowing how users differ from each other ing for Tailoring and Personalizing Interfaces. In Joint Proceedings of the allows us to better alter the interface. These characteristics can ACM IUI 2019 Workshops, Los Angeles, USA, March 20, 2019 , 2 pages. then be used to construct a user model capturing this information. Examples of characteristics that may play a role in how to design an 1 INTRODUCTION optimal interface are cognitive style, personality, and susceptibility When designing interfaces practitioners often rely on knowledge to persuasive strategies. and experience about the interface’s intended users and their needs A second challenge is that of profiling users in terms of these in order to provide the optimal interface for its users. When cre- characteristic based on how they interact with the system. Several ating user interfaces that can be personalized, a more data-driven approaches exist for this more computational challenge, for example approach is mostly taken, where practitioners rely on methods that mining data from social media and clickstream analysis. use implicit or explicit feedback to prescribe how to alter interfaces. A third challenge is knowing how to use knowledge about an ∗ Also affiliated with the BISS Institute, Heerlen, the Netherlands individual user in terms of these characteristics to adapt an interface † Also affiliated with the Dep. of Computer Science, University of Cyprus, Cyprus to match this user. When a user’s characteristics are known, the 1 http://humanize-workshop.org/ interface can be altered to best cater to the user. For example by 2 http://iui.acm.org/2019/ reducing the number of search results for users with a lower need for cognition, or by increasing the diversity of the results for users IUI Workshops’19, March 20, 2019, Los Angeles, USA with a broad taste. Copyright © 2019 for the individual papers by the papers’ authors. Copying permitted These challenges are interconnected and there is no natural order for private and academic purposes. This volume is published and copyrighted by its editors. in which these aspects need to be addressed when personalizing an IUI Workshops’19, March 20, 2019, Los Angeles, USA Graus et al. interface. For example, by analyzing behavior data we can identify in the form of adaptations to the game that allow players to train potential individual characteristics that play a role in people’s needs. themselves to better function in their role. Lee et al. [5] provide an The HUMANIZE workshop provides scholars and practitioners overview of different approaches to learning systems used by Eng- in the field of personalized user interfaces and interactions with lish Language Learners. They propose virtual reality (VR) can ease a venue to discuss and explore the commonalities between the the process of learning a language and acclimating to the culture. sub-problems involved with user interface personalization. An non- In addition, adapting the VR based on the (inferred) personality of exhaustive list of topics for this workshop: the learner may make the learning more efficient. • Identifying models that are (expected to be) useful for personaliz- ing user interfaces (e.g., personality, level of domain knowledge, 3 FORMAT AND CONCLUSION need for cognition, cognitive styles) The workshop focuses on bringing together researchers and profes- • Data mining methods to infer user profiles in terms of cogni- sionals working in the field of Web Adaptation and Personalization, tive/psychological user characteristics from data (e.g., how to User Modeling, Human Factors, User Experience, and Artificial infer personality from social media or domain knowledge from Intelligence, to exchange and share their experiences, new ideas clickstreams) and research results about key aspects (theory, applications and • Theory on how to tailor interfaces to better match certain user tools) of bridging the gap between computational intelligence and profiles (e.g., altering the number of search results, ordering of human intelligence. In this edition, we had the honor to host the interface elements, visual versus textual representations) keynote speech from Dr. Ben Steichen, California State Polytechnic • User studies investigating one or more of the above mentioned University, with title: “So you’ve modeled your user, now what? – Adaptation Techniques for Tailoring and Personalizing Interfaces.” 2 CONTRIBUTIONS Ben talked about a variety of highly sophisticated techniques based A total of six papers was accepted for the third edition of the HU- on both theoretical and statistical models have been developed with MANIZE workshop. Papers were categorized into one of three increased accuracy, as well as breadth of user aspects. On the other topics: 1) mobility, 2) social, and 3) learning. Below is a short de- hand, he argued that this strong focus on modeling typically leaves scription of the topics and the accepted papers: less time and resources for the actual use of these models for the tailoring and personalization of the actual interfaces. In conclu- 2.1 Mobility sion, Steichen presented some of the typical techniques that have Two papers aim to leverage psychological knowledge to influence been used to adapt to users, as well as proposed the use of novel mobility. In their position paper Ferwerda and Lee [1] propose an techniques to make best use of the developed models. app that incorporates psychological concepts such as the need for With HUMANIZE 2019 we hope to have organized another edi- relatedness, to combat the negative effects of physical inactivity. tion in a series of workshops that will address the current challenges Mohan, Klenk, and Bellotti [6] conducted qualitative research to and research directions related to human-centred designs and de- understand what factors play a role in what mode of transporta- velopments, letting the users having always the “final word” in tion people use. The findings of these interviews have been used their interactions with intelligent processes and applications. to design a survey which was distributed and completed by 235 respondents. The survey responses are analyzed and the results REFERENCES show how different factors of the respondents’ geographical situ- [1] Bruce Ferwerda and Michael Lee. 2019. Tamagotchi++: A Serious, Personalized Game to Encourage Healthy Behavior. In Companion Proceedings of the 24th ation, personal situation and personality influences the modes of International conference on Intelligent User Interfaces: 3rd Workshop on Theory- transportation they use. Informed User Modeling for Tailoring and Personalizing Interfaces (HUMANIZE). [2] J. R. Hauser, G. L. Urban, G. Liberali, and M. Braun. 2009. Website Morphing. Marketing Science 28, 2 (mar 2009), 202–223. 2.2 Social [3] Sopan Khosla, Siddhant Arora, Abhilash Nandy, Ankita Saxena, and Anandhavelu Two papers investigate social aspects. Khosla et al. [3] investigated N. 2019. Understanding Community Rivalry on Social Media: A Case Study of Two Footballing Giants. In Companion Proceedings of the 24th International conference hate speech on Twitter in the context of football matches. They col- on Intelligent User Interfaces: 3rd Workshop on Theory-Informed User Modeling for lected and processed data to identify hate speech and subsequently Tailoring and Personalizing Interfaces (HUMANIZE). [4] Domina Kiunsi and Bruce Ferwerda. 2019. Using a Serious Game to Teach User- used this data to see how offline events influence the volume of hate Centered Design. In Companion Proceedings of the 24th International conference speech, how Twitter users that engage in hate speech differ from the on Intelligent User Interfaces: 3rd Workshop on Theory-Informed User Modeling for general Twitter users and the linguistic properties of hate speech. Tailoring and Personalizing Interfaces (HUMANIZE). [5] Michael Lee, Eric Nersesian, Adam Spryszynski, and Ulysee Thompson. 2019. Xu and Lee [7] propose a research direction towards understanding Personalizing VR Educational Tools for English Language Learners. In Companion why current social shopping systems have limited effect on the Proceedings of the 24th International conference on Intelligent User Interfaces: social ties, by investigating how people engage in social shopping, 3rd Workshop on Theory-Informed User Modeling for Tailoring and Personalizing Interfaces (HUMANIZE). their challenges and goals and privacy and risk perceptions. [6] Shiwali Mohan, Matthew Klenk, and Victoria Bellotti. 2019. Exploring How to Personalize Travel Mode Recommendations For Urban Transportation. In 2.3 Learning Companion Proceedings of the 24th International conference on Intelligent User Interfaces: 3rd Workshop on Theory-Informed User Modeling for Tailoring and Personalization could be a means to improve the effectiveness and Personalizing Interfaces (HUMANIZE). [7] Yu Xu and Michael Lee. 2019. Personalizing User Interactions in a Social Shopping efficiency in learning systems. Two papers investigate incorporat- Context and Open Challenges. In Companion Proceedings of the 24th International ing psychological knowledge to do so. Kiunsi and Ferwerda [4] conference on Intelligent User Interfaces: 3rd Workshop on Theory-Informed User propose a serious game to teach user-centered design. While the Modeling for Tailoring and Personalizing Interfaces (HUMANIZE). proposed game itself is already functional, extensions are presented