ACM UMAP – User Modeling, Adaptation and Personalization – is the premier international conference for researchers and practitioners working on systems that adapt to individual users or to groups of users, and that collect, represent, and model user information. ACM UMAP is sponsored by ACM SIGCHI and SIGWEB, and organized with User Modeling Inc. as the core Steering Committee, extended with past years’ chairs. The proceedings are published by ACM and will be part of the ACM Digital Library.
ACM UMAP covers a wide variety of research areas where personalization and adaptation may be applied. This includes a number of domains in which researchers are engendering significant innovations based on advances in user modeling and adaptation, recommender systems, adaptive educational systems, intelligent user interfaces, e-commerce, advertising, digital humanities, social networks, personalized health, entertainment, and many more.
We welcome submissions related to user modeling, personalization and adaptation; the conference web site provides a detailed list. Below we present a short (but not proscriptive) list of topics of importance to the conference. As the theme for UMAP 2021 is “Re-Evaluating Evaluation” we encourage submissions in all areas that offer a critical analysis of evaluations of personalized systems. We particularly want to acknowledge that some of the research might be influenced by Covid-19 related constraints (e.g., difficulty of running lab studies), and welcome submissions which introduce novel methodologies arising from a need to conduct research in new ways.
- Personalized Recommender Systems
- Track chairs: Alejandro Bellogin, Sole Pera, Ludovico Boratto
- Adaptive Hypermedia and the Semantic Web
- Track chairs: Maria Bielikova, Panagiotis Germanakos, Ben Steichen
- Intelligent User Interfaces
- Track chairs: Katrien Verbert, Denis Parra
- Personalized Social Web
- Track chairs: Julita Vassileva, Jie Zhang
- Technology-Enhanced Adaptive Learning
- Track chairs: Ella Haig, Manolis Mavrikis
- Fairness, Transparency, Accountability, and Privacy
- Track chairs: Christine Bauer, Michael Ekstrand
- Personalization for Persuasive and Behavior Change Systems
- Track chairs: Jaap Ham, Rita Orji
- Paper Abstracts: January 17, 2021 (mandatory)
- Full paper: January 24, 2021
- Notification: March 8, 2021
- Camera-ready: April 11, 2021
- Video submission: May 5th, 2021
- Conference: June 21-June 25, 2021 (tentative)
Note: The submissions times are 11:59pm AoE time (Anywhere on Earth)
We welcome submissions related to user modeling, personalization and adaptation in any area. The topics listed below are not intended to limit possible contributions.
Final decisions will be made on the basis of suitability for, and fit to, the overall conference (not for specific tracks). Additionally, there is no quota for the maximal number of accepted papers per track. Topics include (but are not limited to):
Personalized Recommender Systems
Alejandro Bellogin - Universidad Autonoma de Madrid, Sole Pera - Boise State University & Ludovico Boratto - Eurecat
Personalized, computer-generated recommendations have become a pervasive feature of today’s online world. From the traditional book and movie recommendations, to suggestions of what we should eat and wear or where we should travel, recommender systems are seamlessly embedded in our daily lives. The underlying recommender systems are designed to help users and providers in a number of ways. From a user’s viewpoint, these systems assist consumers by identifying relevant and appealing items (e.g., products and services) within large collections. From a provider’s perspective, recommender systems have shown to be valuable tools to steer consumer behavior. Regardless of who are the main stakeholders, the design of recommender systems requires the careful consideration of various aspects, including the choice of the user modeling approach, the recommendation algorithm itself, and the user interface.
This track aims to provide a forum for researchers and practitioners to discuss open challenges, latest solutions and novel research approaches in the field of recommender systems. In addition to mature research works addressing any of the aforementioned technical aspects pertaining to recommendations, we also particularly welcome research contributions that address questions related to the user perception and the business value of recommender systems.
Topics include (but are not limited to):
- Traditional or classic recommendation algorithms
- Context-aware recommender systems (including temporal, social, geographical, etc.)
- User modeling and preference elicitation
- Psychological aspects of recommendation (e.g., psychologically-informed user- and item-modeling and recommendation perception)
- Business value of recommendation systems and multi-stakeholder environments
- Responsible recommendation, including algorithmic bias and fairness, filter bubbles, ethics, and privacy
- Evaluation of personalized recommender systems
- Explanations, transparency, and trust for recommender systems
- Case studies of real-world implementations
- Recommendation in non-traditional domains (e.g., education, health, e-governance, tourism, fashion, energy, etc.)
- Novel recommendation paradigms (e.g., sequence-aware, conversational, harm-aware, etc.)
Adaptive Hypermedia and the Semantic Web
Maria Bielikova - Kempelen Institute of Intelligent Technologies, Panagiotis Germanakos - SAP SE & University of Cyprus & Ben Steichen - California State Polytechnic University, Pomona
Adaptive hypermedia and adaptive web explore alternatives to the traditional “one-size-fits-all” approach in the development of web and hypermedia systems. Typically, these systems build a model of the interests, preferences and knowledge of each individual user, and use this model in order to adapt the behavior of hypermedia and web systems to the needs of that user. However, this area has recently expanded towards complementary research paths in order to improve the interaction with the user. For instance, it has recently acquired methods from the interactive support to information search to enhance the user control during information exploration, with the objective of improving information access by fusing automatic and human capabilities; moreover, it is expanding towards multimodal user interfaces and Augmented Reality to provide users with richer types of content and to support a more engaging fruition of information.
The semantic web frequently serves as an infrastructure to enable adaptive and personalized Web systems. Semantic web technology targets the use of explicit semantics and metadata to help web systems perform the desired functionality: this implies the use of linked data from the web, the use of ontologies in models, or the use of metadata in user interfaces, as well as the use of ontologies for information integration. UMAP aims to provide a forum to researchers to discuss open research problems, solid solutions, latest challenges, novel applications and innovative research approaches in adaptive hypermedia and the semantic web.
Topics include (but are not limited to):
- Web user profiles
- Adaptive navigation support
- Personalized search
- Web content adaptation
- Analytics of web user data
- Adaptive web sites and portals
- Adaptive books and textbooks
- Social navigation and social search
- Navigation support in continuous media and virtual environments
- Usability engineering for adaptive hypermedia and web systems
- Novel methodologies for evaluating adaptive hypermedia and web systems
- Semantic Web technologies for web personalization
- Ontology-based data access and integration/exchange on the adaptive web
- Ontology engineering and ontology patterns for the adaptive web
- Ontology-based user models
- Semantic social network mining, analysis, representation, and management
- Crowdsourcing semantics; methods, dynamics, and challenges
- Semantic web and linked data for adaptation
- Advanced user interfaces, such as VR and AR, for Adaptive Hypermedia
Intelligent User Interfaces
Katrien Verbert - Katholieke Universiteit Leuven & Denis Parra - Pontificia Universidad Catolica de Chile
Intelligent user interfaces aim to improve the interaction between computer systems and human users by means of artificial intelligence.
The systems support and complement different types of abilities that are normally unavailable in the context of human-only cognition. Previous work has found that humans do not always make the best possible decisions when working together with computer systems.
By designing and deploying improved forms of support for interactive collaboration between human decision makers and systems, we can enable decision making processes that better leverage the strengths of both collaborators. More generally this topic can be characterized by exploring how to make the interaction between computers and people smarter and more productive, which may leverage solutions from human-computer interaction, data mining, natural language processing, information visualization, and knowledge representation and reasoning.
Topics include (but are not limited to):
- Adaptive personal virtual assistants (e.g., interaction with social robots)
- Adapting natural interaction (e.g., natural language, speech, gesture)
- Intelligent user interfaces based on sensor data (UIs for cars, fridges, etc.)
- Multi-modal interfaces (speech, gestures, eye gaze, face, physiological info, etc.)
- Intelligent wearable and mobile interfaces
- Smart environments and tangible computing
- Explainable intelligent user interfaces
- Affective and aesthetic interfaces
- Tailored decision support (e.g., over- and under-reliance in uncertain domains)
- Adaptive information visualization
- Scalability of intelligent user interfaces to access huge datasets
- User-centric studies of interactions with intelligent user interfaces
- Novel datasets and use cases for intelligent user interfaces
- Evaluations of intelligent user interfaces
- Interfaces for intelligent visualizations
- Interfaces for personalized and non-personalized recommendation systems
- Implementation and evaluation of Interfaces for Human-centered artificial intelligence
Personalized Social Web
Julita Vassileva - University of Saskatchewan & Jie Zhang - Nanyang Technical University
The social web is continuously growing and social platforms are a fundamental part of our life. Mediated communication is becoming the primary form of communication for young people, and adults follow in increasing numbers. Online communication is increasingly enriched by the use of memes, pictures, audio and video, though language (textual and oral) remains a fundamental tool with which people interact, convey their opinions, construct and determine their social identity. Lifelogging data (e.g., health, fitness, food) is growing as well on the social web. This type of personal information source, gathered for private use through personal devices, is now often shared in online communities. These trends open new challenges for research: how to harness the power of collective intelligence and quantified self data in online social platforms to identify social identities, how to exploit continuous feedback threads, and how to improve the individual user experience on the social web. The social web has many important applications that can benefit society as a whole, for example, mobilizing people to engage in safe behaviours in the pandemic, to support each other in isolation, and live healthier both physically and mentally. Privacy and trust are important issues that arise, since the social web also helps spread misinformation that has harmful effects on individuals and the community. Approaches to deal with these issues are needed. We invite original submissions addressing all aspects of personalization, user models building and personal experience in online social systems.
Topics include (but are not limited to):
- Personalization of the web experience in social systems
- Adaptations based on personality, society, and culture
- Personalization algorithms and protocols inspired by human societies
- Social recommendation
- Identifying social identities in social media
- Social and crowd-generated data for adaptation
- Personalized information retrieval
- Exploiting quantified self data on the social web of things
- Data-driven approaches for personalization
- Modeling individuals, groups, and communities
- Collective intelligence and experience mining
- Pattern and behavior discovery in social network analysis
- Opinion mining for user modeling
- Sentiment analysis
- Topic modeling for online conversations and short texts
- Applications on the social web, e.g. for public safety
- Incentivizing participation & behavior change on the social web
- Privacy, perceived security, and trust and distrust in social systems
- Ethical issues involved in studying the social web
- User awareness and control, social visualization
- Evaluation methodologies for the social web
Technology-Enhanced Adaptive Learning
Ella Haig - University of Portsmouth & Manolis Mavrikis - University College London
Learning is a complex human process that involves cognitive, metacognitive, motivational, affective, and psychomotor aspects that interact with the learning context. Technology is playing an increasing role in the learning process, and technology-enhanced adaptive learning has the potential to support this process in a variety of ways.
By modelling the cognitive, metacognitive, motivational, affective and psychomotor aspects of learning and provide personalized support, technological solutions can improve the effectiveness and efficiency of learning experiences, as well as learner/teacher satisfaction.
Technological innovations bring new opportunities to recognize learners’ needs and to orchestrate suitable learning solutions, with and without the involvement of the teacher. This covers not only formal educational settings, but also lifelong learning requirements (including workplace training) as well as the informal acquisition of skills (e.g., in daily activities, serious games, etc.).
Technology-enhanced adaptive learning is inherently interdisciplinary, covering a wide spectrum of modeling issues and challenges from different perspectives. Therefore, this track invites researchers, developers, and practitioners from various disciplines to present their innovative learning solutions, share acquired experience, and discuss their modeling challenges for personalized adaptive learning.
Topics include (but are not limited to):
- Domain, learner, teacher and context modeling
- Modeling cognitive, metacognitive, motivational, affective, and psychomotor aspects of learning
- Diagnosis of learner needs and calibration of support and feedback
- Adaptive and personalized support for learning
- Dealing with ethical issues involved in detecting and modeling a wider range of information sources (e.g., information from novel sensing devices, ambient intelligent features) that may affect learning
- Management of large, open, and public datasets for educational data mining
- Agent-based learning environments and virtual pedagogical agents
- Open corpus personalized learning
- Collaborative and group learning
- Adaptive technologies to orchestrated classroom Learning
- Personalized teachers’ awareness and support tools
- Multimodal learning analytics to personalize learning
- UMAP aspects in specific learning solutions: educational recommender systems, intelligent tutoring systems, serious games, personal learning environments, MOOCs
- Wearable technologies and augmented reality in adaptive personalized learning
- Processing collected data for UMAP: educational data mining, learning analytics, big data, deep learning.
- Semantic web and ontologies for e-learning
- Interoperability, portability, and scalability issues
- Case studies in real-world educational settings
- New methodologies to develop user-centered highly personalized learning solutions
- Learning analytics
Fairness, Transparency, Accountability, and Privacy
Christine Bauer - Utrecht University & Michael Ekstrand - Boise State University
Adaptive systems researchers and developers have a social responsibility to care about the impact of their technology on individual people (users, providers, and other stakeholders) and on society. This involves building, maintaining, evaluating, and studying adaptive systems that are fair, transparent, respect users’ privacy, and beneficial to society. The importance of these topics is highlighted by their presence in ethical guidelines such as the ACM Code of Ethics (particularly sections 1.2 (Avoid Harm), 1.4 (Be fair), and 1.6 (Respect privacy)), as well as their influence on law and regulation such as the GDPR.
Fairness in modeling and responding to user needs includes the spectrum from algorithmic fairness to social implications of adaptive systems. We define transparency both in terms of system usability and as a means to resolve problems with privacy, fairness, and accountability. Accountability considers mechanisms for identifying violations and enforcing social and other concerns in UMAP applications. Privacy topics include the management of the tradeoff between data collection and user modeling as well as innovative means to resolve privacy problems through algorithms, interfaces, or other technical or non-technical means.
- Bias and discrimination in user modeling
- User, producer, and other stakeholders’ perceptions and expectations of fairness
- Algorithmic methods for measuring and increasing fairness
- UMAP applications for under-served groups
- Balancing needs of users versus system owners and others
- Enhancing/embracing diversity and cultural differences in user modeling
- Ethical concerns in development and study of UMAP applications
- Ethical considerations for user modeling including ‘filter bubble’ or ‘balkanization’ effects
- User perceptions or expectations of transparency in user modeling or personalization
- Algorithmic support for transparency
- Explanations, visualizations, and other interface innovations for transparency
- Explainable and scrutable user modeling
- (User-centric) evaluations of transparency or explanation mechanisms
- Measuring transparency
- Transparency to facilitate accountability and legal compliance
Personalization for Persuasive and Behavior Change Systems
Jaap Ham - Eindhoven University & Rija Orji - Dalhousie University
For solving many different societal problems, changing human behavior is crucial. Persuasive and Behaviour change systems are intentionally designed to support people to achieve behaviour change objectives. Research in the area of Persuasive and Behavior Change Systems has advanced over the years attracting increasing interest from both practitioners and researchers due to the increasing realization of the important role interactive technologies can play in assisting and motivating people to achieve their various goals and objectives. Persuasive and behavior change systems have found applications in various domains including health and wellness, safety and security, environmental sustainability, education, and politics. Research has continuously highlighted the importance of personalizing these systems. Personalization of persuasive and behaviour change systems adapts and tailors these systems to increase their relevance, motivational appeal, user experience, and hence their overall effectiveness at empowering and assisting people to achieve their goals, which are core final aims of user modeling, and adaptation and personalization of systems. Despite the recent advances, research in this area is generally still in infancy and there are many crucial research challenges ready for innovation.
We invite original submissions addressing the broad areas of personalization and tailoring, including but not limited to personalization models, user models, and personal experience designing personalized persuasive and behaviour change systems, computational personalization, design and evaluation methods, and personalized and adaptive behaviour change technologies.
Topics include (but are not limited to):
- Frameworks and models for developing personalized persuasive technology.
- Objective and subjective approaches to personalizing persuasive technologies.
- Methods and Metrics for evaluating the effectiveness of personalized persuasive technology.
- Long-term evaluation and evidence of long-term effects of personalized persuasive technology.
- Methods for large-scale computational personalization of persuasive systems.
- Systematically investigating and highlighting the difference between Adaptivity and Adoptivity.
- Systematically investigating and highlighting the difference between system-controlled personalization and user-controlled personalization.
- The relationships between individual characteristics and effectiveness of various persuasive technology features.
- How to balance the cost and benefit of personalizing persuasive technology.
- How to develop ethical and privacy-sensitive personalized persuasive technology.
- What do we personalize (for example, do we personalize the persuasive strategies, approaches, or end-goals)?
- How do we personalize (e.g., subjective and objective personalization methods)?
- Who do we personalize for (e.g., personality, gender, age, persuadability, player types, emotional states, contextual/situational variables)?
- Challenges and limitations of implementing personalized persuasive technology and possible solutions.
- Case studies and examples of personalized persuasive technologies.
- Success and failure stories with regard to personalized persuasive technology.
Submission and Review Process
Papers will be submitted through EasyChair:
Long (8 pages + references) and Short (4 pages + references) papers in ACM style. Original research papers addressing the theory and/or practice of UMAP, and papers showcasing innovative use of UMAP and exploring the benefits and challenges of applying UMAP technology in real-life applications and contexts are welcome.
- Long papers should present original reports of substantive new research techniques, findings, and applications of UMAP. They should place the work within the field and clearly indicate its innovative aspects. Research procedures and technical methods should be presented in sufficient detail to ensure scrutiny and reproducibility. Results should be clearly communicated and implications of the contributions/findings for UMAP and beyond should be explicitly discussed.
- Short papers should present original and highly promising research or applications. Merit will be assessed in terms of originality and importance rather than maturity, extensive technical validation, and user studies. Separation of long and short papers will be strictly enforced so papers will not compete across categories, but only within each category.
Papers must be formatted using the ACM SIG Standard (SIGCONF) proceedings template: https://www.acm.org/publications/proceedings-template.
UMAP uses a double blind review process. Authors must omit their names and affiliations from submissions, and avoid obvious identifying statements. For instance, citations to the authors' own prior work should be made in the third-person.
All accepted papers will be published by ACM and will be available via the ACM Digital Library. To be included in the Proceedings, at least one author of each accepted paper must register for the conference and present the paper there. Student registration fee is allowed for students who present a student paper.
- Nava Tintarev, University of Maastricht, the Netherlands
- Marko Tkalcic, University of Primorska, Slovenia