From Modeling and Profiling to Auditing:

the Evolving Role of the User in Recommender System Research

Held as part of the 34th ACM Conference on User Modeling, Adaptation and Personalization (UMAP 2026),
Gothenburg, Sweden, June 8 — 11, 2026

* Banner Image generated by Gemini 3 Flash

About

Recommender Systems (RSs) are central to how users interact with online platforms, from streaming services to e-commerce and social media, and understanding users is essential not only to improve their overall experience but also to identify, analyze, and mitigate harmful system behaviors. At the heart of recommender systems lies user modeling and profiling, which capture behavioral patterns, preferences, and contextual information to personalize content and enhance user experience. Understanding how users are represented is crucial not only for designing effective recommendations but also for evaluating the broader societal and ethical implications of these systems.

Importantly, this need has become increasingly salient in light of recent regulatory developments, such as the European Union’s Digital Services Act (DSA) and AI Act, which explicitly call for greater transparency, risk assessment, and accountability of algorithmic systems, including RSs. These regulations elevate auditing from a research-driven activity to a regulatory and societal requirement, reinforcing the need for systematic methods to examine how RSs operate in practice.

Under this lens, users themselves can play an active role in assessing harmful system behavior. Algorithmic auditing has emerged as a key approach to investigate how RSs function in practice, uncovering biases, unintended behaviors, and potential harms. By connecting insights from user modeling to auditing practices, researchers and practitioners can move from passively representing users to actively characterizing them as participants in the assessment process. This perspective emphasizes the sociotechnical nature of recommender systems, highlighting the interplay between algorithms, data, and human interactions.

The proposed tutorial provides a comprehensive exploration of this continuum, from modeling and profiling to auditing, guiding participants through the historical foundations, evolving concepts, and recent paradigm shifts in user representation, while demonstrating how these insights inform practical, participatory auditing strategies. Participants will leave with a comprehensive understanding of how users can be both modeled and actively involved in shaping accountable, transparent, and human-centered RSs.

Learning objectives

By the end of the tutorial, participants will be able to:

Trace the historical development of user modeling and profiling in recommender systems research.
Describe evolving methodologies and paradigms in user modeling and profiling.
Identify the motivations for algorithmic auditing of recommender systems.
Recognize the various ways users can participate in auditing processes.
Apply insights from user modeling and profiling to the design and execution of effective audits.

Targeted audience

The tutorial is intended for researchers, practitioners, and professionals within the UMAP community, as well as those exploring the broader field of user modeling, profiling and/or auditing, with a specific focus on recommender system research. It accommodates participants with varying levels of experience, from beginners to experts, and requires no specific prerequisites. The content is structured to be accessible to a diverse audience, and the concepts and applications discussed span multiple disciplines, including human-computer interaction, social computing, recommender systems, and science for policy, making the session relevant to an interdisciplinary audience.

Structure

The proposed tutorial provides a comprehensive exploration of the user modeling landscape and its connection to algorithmic auditing.

The following outline summarizes its structure:

Opening and welcome
User modeling and profiling
1. Historical foundations: Reviews the origins and key milestones of user modeling and profiling research, including early methods for capturing user behavior and preferences.
2. Terminology and conceptual landscape: Examines key terms and evolving concepts in user modeling and profiling, highlighting the crucial implications for designing, interpreting, and applying user representations.
3. Recent paradigm shifts: Explores the evolution from explicit to implicit and pseudo-explicit profiling, and advances in user behavior modeling and universal representations.
User involvement in algorithmic auditing
1. Introduction to auditing: Presents the role of algorithmic auditing in recommender systems, emphasizing transparency, accountability, and the socio-technical context.
2. Modes of user participation: Reviews user-initiated, user-driven, and user-engaged audits, showing how users may contribute to identifying, reporting, and mitigating harmful system behaviors.
3. Case studies and challenges: Illustrates examples of user involvement in recommender audits and discusses practical opportunities and challenges for participatory approaches in real-world systems.
Conclusions
1. From modeling and profiling to auditing: Connects insights from user modeling and profiling to algorithmic auditing, highlighting emerging practices exploring participatory, community-centered approaches for accountable recommender systems.
2. Q&A and discussion: Opens the floor for participant questions and discussion, summarizing key takeaways and actionable insights.

Presenters

Lorenzo Porcaro

He is a Marie Skłodowska-Curie Postdoctoral Fellow at the Department of Computer, Control and Management Engineering (DIAG) of the Sapienza University of Rome (Italy), where he is PI of the project titled Algorithmic Auditing for Music Discoverability (AA4MD). Before, he served as Scientific Project Officer at the European Centre for Algorithmic Transparency (ECAT) of the European Commission's Joint Research Centre (JRC). His work focuses on assessing the impact of recommender systems on their users. He holds a PhD in Information and Communication Technologies from Universitat Pompeu Fabra (Spain). His research interests include recommender systems, algorithmic auditing, and human-computer interaction.

Erasmo Purificato

He has been a Scientific Project Officer at the European Centre for Algorithmic Transparency (ECAT) of the European Commission's Joint Research Centre (JRC) since 2024. Previously, he was a Research Assistant in the “Human-Centred AI” group at Otto von Guericke University Magdeburg (OVGU) and in the “Human-Centred Technologies for Educational Media” department at the Leibniz Institute for Educational Media | Georg Eckert Institute (GEI).

He holds a PhD in Computer Science from OVGU with a dissertation entitled “Fairness Analysis of Graph Neural Networks for Behavioral User Modeling”, which won the Runner-up prize of the 2025 Informatics Europe Best Dissertation Award. Beyond the mentioned tutorials, Erasmo co-organized several workshops at ECIR, IUI, and CHItaly. He has been part of the organizing and program committee of various conferences and workshops, such as SIGIR, RecSys, UMAP, HT, and ExUM, and served as a reviewer for many journals such as IPM, MTAP, TORS, and TITS. Among the awards, he won the Best Full Paper Award at RecSys ’25 with the article “You Don't Bring Me Flowers: Mitigating Unwanted Recommendation with Conformal Risk Control”.

Registration

Registration to the tutorial will be managed by the UMAP 2026 main conference organization.

Contacts

Lorenzo Porcaro: lorenzo [dot] porcaro [at] uniroma1 [dot] it

Erasmo Purificato: erasmo [dot] purificato [dot] acm [dot] org

Acknowledgments

This tutorial is part of the project Algorithmic Auditing for Music Discoverability (AA4MD) which has received funding from the European Union’s Horizon Europe research and innovation programme under the Marie Skłodowska-Curie grant agreement No 101148443. It is also partially supported by the HUMAINT programme (Human Behaviour and Machine Intelligence), Joint Research Centre, European Commission.

References

Alicia DeVos, Aditi Dhabalia, Hong Shen, Kenneth Holstein, and Motahhare Eslami. 2022. Toward User-Driven Algorithm Auditing: Investigating Users’ Strategies for Uncovering Harmful Algorithmic Behavior. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems (CHI ’22), New Orleans, LA, USA. Association for Computing Machinery, New York, NY, USA, Article 626, 19 pages. https://doi.org/10.1145/3491102.3517441

Christopher Ifeanyi Eke, Azah Anir Norman, Liyana Shuib, and Henry Friday Nweke. 2019. A Survey of User Profiling: State-of-the-Art, Challenges, and Solutions. IEEE Access 7 (2019), 144907–144924. https://doi.org/10.1109/ACCESS.2019.2944243

Michael D. Ekstrand, Afsaneh Razi, Aleksandra Sarcevic, Maria Soledad Pera, Robin Burke, and Katherine Landau Wright. 2025. Recommending With, Not For: Co-Designing Recommender Systems for Social Good. ACM Transactions on Recommender Systems (August 2025). https://doi.org/10.1145/3759261 (Just Accepted).

European Commission. 2022. Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Single Market for Digital Services and amending Directive 2000/31/EC (Digital Services Act). Official Journal of the European Union. https://eur-lex.europa.eu/eli/reg/2022/2065/oj

European Commission. 2024. Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 on Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act). Official Journal of the European Union. https://eur-lex.europa.eu/eli/reg/2024/1689/oj

Sumitkumar Kanoje, Sheetal Girase, and Debajyoti Mukhopadhyay. 2015. User Profiling Trends, Techniques and Applications. https://doi.org/10.48550/arXiv.1503.07474

Joseph A. Konstan and Loren G. Terveen. 2021. Human-Centered Recommender Systems: Origins, Advances, Challenges, and Opportunities. AI Magazine 42, 3 (2021), 31–42. https://doi.org/10.1609/aimag.v42i3.18142

Rena Li, Sara Kingsley, Chelsea Fan, Proteeti Sinha, Nora Wai, Jaimie Lee, Hong Shen, Motahhare Eslami, and Jason Hong. 2023. Participation and Division of Labor in User-Driven Algorithm Audits: How Do Everyday Users Work Together to Surface Algorithmic Harms? In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (CHI ’23), Hamburg, Germany. Association for Computing Machinery, New York, NY, USA, Article 567, 19 pages. https://doi.org/10.1145/3544548.3582074

Lorenzo Porcaro, Agnese Macori, Daniel Raffini, and Tiziana Catarci. 2025. User Involvement in Recommender System Audits: A Systematic Review. https://doi.org/10.5281/zenodo.17641849

Erasmo Purificato, Ludovico Boratto, and Ernesto William De Luca. 2024. User Modeling and User Profiling: A Comprehensive Survey. https://doi.org/10.48550/arXiv.2402.09660

Markus Schedl, Vito Walter Anelli, and Elisabeth Lex. 2025. Technical and Regulatory Perspectives on Information Retrieval and Recommender Systems: Fairness, Transparency, and Privacy. The Information Retrieval Series, Vol. 50. Springer, Cham. https://doi.org/10.1007/978-3-031-69978-8

Hong Shen, Alicia DeVos, Motahhare Eslami, and Kenneth Holstein. 2021. Everyday Algorithm Auditing: Understanding the Power of Everyday Users in Surfacing Harmful Algorithmic Behaviors. Proceedings of the ACM on Human-Computer Interaction 5, CSCW2, Article 433 (October 2021), 29 pages. https://doi.org/10.1145/3479577

Jonathan Stray, Alon Halevy, Parisa Assar, Dylan Hadfield-Menell, Craig Boutilier, Amar Ashar, Chloe Bakalar, Lex Beattie, Michael Ekstrand, Claire Leibowicz, Connie Moon Sehat, Sara Johansen, Lianne Kerlin, David Vickrey, Spandana Singh, Sanne Vrijenhoek, Amy Zhang, McKane Andrus, Natali Helberger, Polina Proutskova, Tanushree Mitra, and Nina Vasan. 2024. Building Human Values into Recommender Systems: An Interdisciplinary Synthesis. ACM Transactions on Recommender Systems 2, 3, Article 20 (June 2024), 57 pages. https://doi.org/10.1145/3632297

Page updated

Google Sites

Report abuse