< Back Ulysse Maes PhD Researcher ulysse.jan.l.maes@vub.be Ulysse joined SMIT in September 2023 as a PhD researcher. His focus is on comprehending the multifaceted challenges and opportunities presented by the evolving landscape of AI, particularly in news and video recommender systems. He holds a master’s degree in Business and Information Systems Engineering, where he graduated magna cum laude. His academic pursuits took him to SKEMA Business School in Antibes, where, during an Erasmus exchange, he deepened his knowledge on entrepreneurship in an international setting. Notably, his master’s thesis was dedicated to the development of a user-oriented chatbot enhancing DMN execution. This was followed by his successful completion of a Postgraduate Degree in Tech Innovations in Ventures and Teams. During that year, he co-founded Amodamo , a platform endorsing responsible fashion, and established Scribewave , an online transcription service specialised in converting speech-to-text in local languages. Visit my research profile

< Back 12 Feb 2024 Reflections on the research presented at Etmaal On Thursday 8th and Friday 9th February 2024, Hanne Vandenbroucke and Ulysse Maes attended the 26th edition of Etmaal van Communicatiewetenschap in Rotterdam. At this two-day conference, researchers from Communication Sciences in Belgium and The Netherlands unite to present their projects, get feedback from peers, and be inspired by the work of others. They were invited to present their work-in-progress during the Research Escalator. In this blogpost we highlight the key take-aways from their presentations. Multi-stakeholder approach to news personalization - Hanne Vandenbroucke What lies behind the "For you", "Read more" or "See also" sections on the website or mobile app of your favorite news brand? By conducting stakeholder interviews with professionals working in commercial news organizations operating in Flanders: DPG Media, Mediahuis and Mediafin; We aim to map the development and implementation of recommender systems. The key internal stakeholder groups involved in and impacted by news recommender systems are: (1) the newsroom, (2) the technical development, and (3) the commercial business unit. Based on the stakeholder interviews, we are able to build upon the multi-stakeholder framework of Smets et al. (2022) . The preliminary results give insights in the actual decision-making process for recommender development. News organisations started experimenting with a news recommender system on average 3 years ago. The initial process of trial and error had transformed into an ongoing cycle of adjusting the RS design. In practice both the newsroom and the business unit express their objectives, preconditions and concerns to the product owner who aligns the different perspectives and formulates a concrete set of goals. Next, the technology development team will operationalize these objectives into computational metrics and adjust the recommender system design. Performance data is continuously being collected and monitored. A feedback loop set up to communicate the results of the adjusted recommender to the product owner who – together with the data analytics team – derives insights from the data and report back to the business unit and newsrooms. Exploring the influence of misleading explanations on the perceived quality of recommender systems – Ulysse Maes Nowadays, recommender systems are everywhere: you find them on Amazon, on Netflix and Spotify, for example. These algorithmic curation systems help internet users to efficiently navigate through vast amounts of content. While holding clear advantages in terms of user experience, there are some limitations and normative concerns. One of these concerns stems from the limited transparency they provide. This may lead to distrust and frustration. Adding explanations may yield different results, which depend on your objectives. (Tintarev & Masthoff, 2007) suggest seven goals of explainable recommendations: effectiveness, efficiency, satisfaction, scrutability, transparency, trust, and persuasiveness. Note that maximizing for one goal might be beneficial for another (e.g. improving for scrutability- giving users the ability to change the outcomes to their liking- might improve satisfaction). However, optimizing for one goal might also harm other goals. This research specificially dives into the possible conflict between optimizing for persuasiveness and its effects on transparency and trust. By “optimizing for persuasiveness”, we mean creating compelling narratives to persuade users to consume recommended content. For example, when you buy a pair of jeans online, the shop can try to upsell by recommending some white t-shirts and explain the recommendations as “Style advice for the perfect shirt to wear on your new jeans.” A more neutral explanation could be: “Customers also bought.” Persuasion in itself is not problematic, but it might become problematic once it becomes misleading: hiding important information or even lying about the working of the system, or the drivers behind a decision. As explanations are often linked to a conception of transparency, both with end-users and academia, the mere presence of explanations can already lead to increased trust in the system. By crafting compelling but incorrect explanations, it might be possible to manipulate users into consuming certain content, while still giving them a feeling of agency. There are clearly some incentives to create misleading explanations. But do they work? Theory is still inconclusive. While some research mentions the effectiveness of personalized persuasion (Burtell & Woodside, 2023), others highlight the detrimental effects on long-term trust: would platforms really jeopardize this? Another argument against the effectiveness of misleading explanations is that the costs don’t outweigh the benefits – a critique sometimes formulated on explanations in general as well. The field of explainable AI recognizes the immense potential of Large Language Models (LLMs), such as ChatGPT, to generate personalized, dynamic explanations at scale. It has already been shown that these explanations can be more persuasive than human-generated texts. One of the reasons why we do not see LLM-generated explanations popping up everywhere, is because of their tendency to “hallucinate” - to make up plausible, but incorrect narratives. Previous Next A glimpse into our work-in-progress

< Back 5 Jun 2025 Our presentations at the emma Conference 2025 On the first day of the EMMA Conference, Aina, Hanne, and Pieter joined over 20 other PhD students from all over Europe for the inspiring PhD Workshop. In small groups, under the guidance of an experienced researcher, each student got the opportunity to present some issues related to their PhD track, delving into theoretical frameworks, methodological approaches, data analysis, the writing process, and mental challenges such as dealing with impostor syndrome, The diversity of research themes and perspectives is of great value to our three PhD’ers as their starting the third year of their track. The workshop provided room for deepening, exchange, and building a valuable network of fellow researchers. And of course, there was also time to enjoy the sunshine, have an espresso, Aperol Spritz, and taste some Italian delicacies! 😋🇮🇹🍕 The second day of the EMMA Conference was dominated by fascinating presentations, including that of our colleague Aina , who presented her research with great flair. Aina’s ongoing research (as part of the ALGEPI project ) investigates how Europe’s AI regulatory frameworks—ranging from the EU’s AI Act and European Media Freedom Act (EMFA) to UNESCO and Council of Europe guidelines and national AI strategies in Belgium, France, and Spain—are or are not reshaping newsroom governance. The goal of this research is to explore the intersection between regulation and journalism more closely trying to understand how AI regulation translates into newsroom realities across European contexts. Through a two‐step approach of document analysis and interviews with legal experts, media managers, and practitioners, she discussed the impact of both hard and soft law in the journalistic context as well as media‐tailored AI policies that balance compliance with editorial independence. Next to Aina’s insights, it was inspiring to hear from other research themes and get some findings from other studies. After a strong day at the conference, we had a chance to explore beautiful Rome and do some sightseeing. The day ended in style with a delicious buffet full of authentic Italian dishes.🍝🍦 On day three of the EMMA Conference, it was Hanne and Pieter's turn to present their research. Together, they gave an engaging joint presentation in which they shared a comparative analysis of how decision-making around recommender systems for media distribution is shaped in commercial media companies and public broadcasters. Building upon the framework of Smets et al (2022), they investigate how the strategic perspective on integrating recommender systems differs in profit-oriented organizations and PSM. By integrating their individual case studies, they offered valuable insights into the divergent strategic goals, recommendation purposes, and the implications for the recommender system design. The take-away? There are some similar strategic goals and recommendation purposes for editorial, commercial, and technological stakeholders in both types of media organizations. Nevertheless, the interpretation of concepts such as 'relevant' and 'engagement' differs based on the organizational cultures in profit-driven and public service media. Slide of presentation: Preliminary findings on shared overarching strategic goals and recommendation purposes For those who cannot get enough of Pieter , he also gave an individual presentation afterwards in which he elaborated on his case study work on Flemish public broadcaster VRT. With a focus on the tensions and possible conflicts that arise between stakeholders in the decision-making processes that shape the recommender system development, his presentation elaborated on the stakeholders involved at PSM and the implications of (dis) alignment of their strategic interests. His VRT deep dive hinted at the merits of a stronger involvement of indirect stakeholders for public service delivery and provided learnings on how PSM could tackle future structural integrations of technological innovations, for example, that of generative AI, which was a heavily debated topic at this conference. Previous Next Ciao from EMMA in Rome!

< Back Streaming Affordances for small media markets Living Lab 2023-2026 Delayed viewing, major ad giants like Google and Facebook, and the rise of American streaming platforms have significantly increased the pressure on local providers. In response, Flemish broadcasters and distribution entities have launched their own streaming services in recent years. Some offer free services, with or without ads (such as VRT MAX, VTM GO, and GoPlay), while others operate on subscription models (like Streamz). The local platforms often adopt the successful features of major streaming services. It is crucial for local providers to differentiate themselves with a wealth of local content, giving them a competitive edge over foreign platforms. However, a key challenge in our market is the limited scale, which necessitates strategic choices in purchasing, production, distribution, and marketing. At the heart of these changes is a shift in our viewing habits. To give Flemish streaming services the best chance in this highly competitive market, it is essential to understand how Flemish consumers watch video content, identify key viewing patterns, know what viewers expect from streaming services, and gauge the effectiveness of brand or content positioning strategies. Our living lab, ‘Streaming Affordances for Small Media Markets,’ seeks to address this through five key deliverables: a (1) SOTA analysis, (2) diary study, (3) viewing survey, (4) a living lab, and (5) experiments. 1. SOTA analysis The SOTA analysis compiles data from recent studies and reports, aiming to provide a clear view of trends impacting our audiovisual sector. It covers international, European, and Flemish media usage, alongside 10 trends in the global streaming market that are likely to influence Flanders. 10 trends in de internationale streamingmarkt .pdf Download PDF • 2.21MB The policy brief (in Dutch) 2. Diary study This deliverable explores the motivations that guide Flemish viewers in their selection of VoD services and content. It draws on a large-scale multifaceted audience study with VoD consumers in Flanders and consists of an exploratory survey (N=77), a diary study (N=62) and semi-structured face-to-face interviews (N=60). Motivaties voor VoD-consumptie in Vlaanderen .pdf Download PDF • 1.26MB 3. Viewing survey Building on the diary study, the SMIT viewing survey (N=2000) serves as larger-scale quantitative research. It explores viewers’ attitudes toward linear TV, their familiarity with and use of VoD services, associations with specific platforms, device preferences, and the dynamic between cinema and streaming. Voorkeuren in VoD-consumptie in Vlaanderen. Resultaten kwantitatieve bevraging. .pdf Download PDF • 1.37MB The policy brief on the diary study and the viewing survey (in Dutch) Whitepaper: Finding and discovering in streaming interfaces This whitepaper discusses the role of interface design in supporting and balancing findability and discoverability in line with the literature on HCI and RecSys, and industry reports. This report is not intended to be an exhaustive overview of the existing literature or current practice. It offers a concise synthesis of recent insights into how interface design in streaming services affects the findability and discoverability of content, intended to inspire and highlight the importance of interface design by bringing together academic research and industry reports. We hope this will contribute to a richer dialogue between research and practice, aimed at helping researchers and professionals develop interfaces that support users in finding and discovering content. NL - Whitepaper Streaming Interfaces .pdf Download PDF • 1.78MB EN - Whitepaper Streaming Interfaces .pdf Download PDF • 12.69MB Streaming Europe: How Transnational Streaming is Reshaping European Film and Television Markets Read Researchers on this project Wendy Van den Broeck Associate professor Tim Raats Associate Professor Annelien Smets Research Professor Hanne Vandenbroucke PhD Researcher Consortium partners VLAIO.png VLAIO.png 1/1 Tags Platforms, Public Service Broadcasting, Streaming, Recommenders Previous Next

< Back LLMs and Recommenders Ulysse Maes PhD Researcher My four-year PhD investigates how large language models are reshaping recommender and content-discovery interfaces. Rather than focussing on the underlying recommendation algorithms, I'm exploring how the interfaces around them can contributes. The aim throughout is to bridge new possibilities with generative AI with genuinely user-centric design. My central research question focuses on composing an interface (its layout, its components, what it foregrounds) to match a reader's actual, evolving information need, and proving that this match matters. The first phase of my research will map where LLMs can most usefully shape recommender and content interfaces, and just as importantly, where they shouldn't. That work will produce a validated taxonomy of what readers actually need from a complex, evolving story (a quick check-up versus broad exploration, for instance). I love building prototypes for mobile news reading, that compose interface components (timelines, actor maps, comparison views) to match the user need rather than defaulting to one layout for everyone. Some pressing ethical questions are importznt here: the guiding measure is whether the interface helps someone actually explore and understand — not how long it keeps them engaged. I'm also exploring how to turn tat taxonomy into an evaluation: a controlled comparison of interfaces that match a reader's need against interfaces that don't, alongside a generic relevance-only baseline and the plain article as it exists today. Recommender systems are usually evaluated by whether people click, not by whether the presentation actually served what they came for — this phase is built to answer the second question directly, on mobile, where interface real estate is scarce and the design trade-offs are sharper due to limited screen size. Positioned at the intersection of two research lines — the long-running academic tradition of adaptive interfaces and the fast-moving, industry-driven push toward generative UI — the project's contribution is the piece the second line is currently missing: a validated theory of which interface structures actually serve which reader needs. The goal: making sure that as recommender and content interfaces get rebuilt around generative AI, they stay accountable to the people using them, not just to engagement metrics. Publications Towards Adaptive Mixed-Initiative Media Recommender Systems with LLMs 22/03/2026 FIRE: A Modular Framework for Prototyping and Evaluating Interactive Recommender Explanations 07/06/2026 Let Me Introduce You: Stimulating Taste-Broadening Serendipity through Song Introductions 08/06/2026 Not One News Recommender To Fit Them All: How Different Recommender Strategies Serve Various User Segments 22/09/2025 GenUI(ne) CRS: UI Elements and Retrieval-Augmented Generation in Conversational Recommender Systems with LLMs 08/10/2024 Supervisors Annelien Smets Research Professor Tim Raats Associate Professor Related news 2 Impressions of IBC 2025 Ulysse’s Impressions of the International Broadcasting Convention 2025: Shaping The Future. 2 Looking back at IUI 2025 Ulysse attended the conference and presented at the AXAI workshop 2 Looking back at RecSys 2024 Sharing our reflections on the ACM RecSys Conference 2024 Previous Next

< Back 28 Oct 2024 Throwback on the international seminar on diversity and VOD services Tim Raats and Catalina Iordache participated in the international seminar on diversity and Video-on-Demand (VOD) services in Madrid, hosted by the Diversidad Audiovisual research group at University Carlos III Madrid. This event brought together scholars, industry experts, and policymakers to discuss the challenges and opportunities for promoting diverse content in the increasingly globalized VOD landscape. Tim Raats contributed to the discussion on quotas for European works, presenting on October 24 during a panel focused on promoting European and Spanish content. His talk examined the importance of quota regulations and content prominence to ensure visibility for European productions, underscoring how these measures can support local cultural diversity on international streaming platforms. Catalina Iordache also presented on the same day, addressing the financial impact of foreign streaming platforms on European fiction series. Speaking in a panel on investment and financing of European works, she shared insights into how international players like Netflix and Amazon Prime are contributing to and reshaping the European audiovisual ecosystem by funding local productions. The seminar recordings and a full list of panels are available on the Diversidad Audiovisual website , offering a wealth of information on the dynamics of diversity in the digital streaming era. Previous Next Diversity in VOD services: Seminar highlights from Madrid

< Back Let Me Introduce You: Stimulating Taste-Broadening Serendipity through Song Introductions Brett Binst, Ulysse Maes, Martijn Willemsen, and Annelien Smets. 2026. Let Me Introduce You: Stimulating Taste-Broadening Serendipity through Song Introductions. In 34th ACM Conference on User Modeling, Adaptation and Personalization (UMAP ’26), June 08–11, 2026, Gothenburg, Sweden. ACM, New York, NY, USA, 10 pages. https://doi.org/10.1145/3774935.3806164 Research on how people experience music emphasizes the importance of exploration and diversity in listening. However, music recommender systems struggle with facilitating exploration. Even when music recommender systems are able to recommend something valuable to users that is outside their typical preferences, it still remains difficult to spark their interest. This paper presents a user study examining the efficacy of immersive and informative introductions in stimulating interest in songs that are beyond one’s usual preferences, an experience called Taste-Broadening Serendipity. We uncover two important mechanisms behind the effect of introductions: transportation and cognitive elaboration. Our findings indicate that transportation (i.e., being absorbed into a narrative world) is the strongest predictor of Taste-Broadening Serendipity, while cognitive elaboration (i.e., learning something new about the artist or social context in which the music emerged) has a weaker effect but is easier to stimulate. We propose that song introductions can play an important role in facilitating exploration and increasing diversity of listening on music streaming platforms. Previous Read the article Next

< Back Challenges and opportunities for recommender systems in media markets Ranaivoson, H., Smets, A. & Ballon, P. (2024). Challenges and opportunities for recommender systems in media markets. In U. Rohn, M. Rimscha & T. Raats (Ed.), De Gruyter Handbook of Media Economics (pp. 215-228). Berlin, Boston: De Gruyter. https://doi.org/10.1515/9783110793444-015 Recommender systems have become omnipresent in media markets, fromproduction to distribution and consumption. What is often overlooked, however, isthat the multi-sided nature of contemporary media markets implies that recom-mender systems need to balance the needs and requirements of actors on differentsides of these markets, such as content providers and end-users. Starting from an overview of the rapidly growing literature on recommender systems in media mar-kets, and emphasizing their multi-sided nature, we first describe the main typologies of recommender systems and analyze how various stakeholders might adopt particular strategies to have these systems create novel market opportunities. Then, we discuss media market dilemmas that result from these stakeholders’ often competing objectives in terms of biases, notably filter bubbles, management dilemmas, and is-sues around privacy and transparency. We argue how revisiting these concerns from a multi-stakeholder perspective opens up new angles to understand recommender systems’ impact on contemporary media markets. Recommender Systems Previous Read the article Next

< Back Publishing public service media on demand: A comparative study of public service media companies’ editorial practices on their VoD services in the age of platformization Bruun, H., Johnson, C., Lassen, J. M., Nucci, A., Raats, T., & Świtkowski, F. (2025). Publishing public service media on demand: A comparative study of public service media companies’ editorial practices on their VoD services in the age of platformization. Journal of Digital Media & Policy . This article contributes to the emerging empirical research on the editorial practices of video-on-demand (VoD) publishing in European public service media (PSM). It presents results from a comparative study of the editorial practices visible on the VoDs from ten PSM companies across six countries: United Kingdom, Belgium, Poland, Italy, Canada and Denmark. The aim of the article is to map and compare the editorial practices in the ‘prime space’ of the VoD services and the ‘prime time’ of the companies’ main linear channels. The analysis is based on data from a sample week of 13–19 November 2023. The article contributes to research addressing the key issue of universality in terms of content and discusses the conceptualizations of the audience that seem to be at work in the transition towards an online PSM identity. Public Service Media, Platforms Previous Read the article Next

< Back Welcome to the Metrics Jungle: Organizational Stakeholder Perspectives on Evaluation of News Recommender Systems in Industry Vandenbroucke, H., Michiels, L., & Smets, A. (2025). Welcome to the Metrics Jungle: Organizational Stakeholder Perspectives on evaluation of news recommender systems in industry. ACM Transactions on Recommender Systems, Highlights of RecSys ’24. https://doi.org/10.1145/3778173 News recommender systems (NRS) are increasingly implemented in news industry to enhance digital distribution, yet their development and evaluation are shaped by a complex interplay of organizational stakeholders with diverse objectives. This study investigates the real-world complexities of NRS implementation, considering the multitude of objectives on the NRS provider side by exploring how different stakeholder groups —from journalistic, commercial and tech logic— conceptualize NRS goals and assess their performance. Through 22 semi-structured expert interviews across four news organizations in Europe, we uncover a `jungle of metrics' where an overabundance of data insights does not necessarily lead to better decision-making. While traditional accuracy-based metrics like click-through rate (CTR) remain widely used, they fail to capture the broader strategic objectives that guide NRS adoption in industry. This paper identifies a critical gap between current academic evaluations of NRS and the business key performance indicators (KPIs) used in practice. Rather than evaluating NRS in isolation, stakeholders assess their contribution to overarching organizational goals, including reach, engagement, conversion, and retention. Our findings highlight the need for a more nuanced, multi-stakeholder approach to NRS evaluation that aligns with both business imperatives and journalistic values. Newsmedia, Recommender Systems, Stakeholders Previous Read the article Next

< Back Piecing the puzzle of VOD offerings: A comparison of eight US and local services in Flanders Iordache, C., Raats, T., Pakvis, M., & Asmar, A. (2023). Piecing the puzzle of VOD offerings: A comparison of eight US and local services in Flanders. Journal of Digital Media and Policy , 14 (3), 357 - 376. https://doi.org/10.1386/jdmp_00134_1 The study conducts a catalogue analysis that zooms in on the small market of Flanders (Belgium), and the content strategies of the eight most popular local and US-based players in the market. Through the comparative analysis, we investigate three main directions. First, we analyse the similarities and differences in the catalogue offerings of US and local video-on-demand (VoD) services. Second, we delve into the strategies of local players as they compete with US streamers and other legacy players. Third, we analyse the share and the characteristics of the Flemish titles on offer. The data was collected in the spring of 2022 and consists of 13,500 title entries. The findings point to a set of complex market dynamics. Local services and US-based players are competing and collaborating to create a jigsaw puzzle of on-demand offerings, through various strategies of complacency, resistance, differentiation and diversification/mimicry. Platforms Previous Read the article Next

< Back What is Serendipity? An Interview Study to Conceptualize Experienced Serendipity in Recommender Systems Binst, B. , Michiels, L. & Smets, A. , 2025, (Accepted/In press).What is Serendipity? An Interview Study to Conceptualize Experienced Serendipity in Recommender Systems. Proceedings of the 33rd ACM Conference on User Modeling, Adaptation and Personalization Serendipity has been associated with numerous benefits in the context of recommender systems, e.g., increased user satisfaction and consumption of long-tail items. Despite this, serendipity in the context of recommender systems has thus far remained conceptually ambiguous. This conceptual ambiguity has led to inconsistent operationalizations between studies, making it difficult to compare and synthesize findings. In this paper, we conceptualize the user’s experience of serendipity. To this effect, we interviewed 17 participants and analyzed the data following the grounded theory paradigm. Based on these interviews, we conceptualize experienced serendipity as a user experience in which a user unintentionally encounters content that feels fortuitous, refreshing, and enriching. We find that all three components—fortuitous, refreshing and enriching—are necessary and together are sufficient to classify a user’s experience as serendipitous. However, these components can be satisfied through a variety of conditions. Our conceptualization unifies previous definitions of serendipity within a single framework, resolving inconsistencies by identifying distinct flavors of serendipity. It highlights underexposed flavors, offering new insights into how users experience serendipity in the context of recommender systems. By clarifying the components and conditions of experienced serendipity in recommender systems, this work can guide the design of recommender systems that stimulate experienced serendipity in their users, and lays the groundwork for developing a standardized operationalization of experienced serendipity in its many flavors, enabling more consistent and comparable evaluations. Recommender Systems Previous Read the article Next