< Back 16 Oct 2024 A new website with all there is to know about Video-on-Demand (VoD) research methods Exciting news for all media researchers working on VOD: a new website has been launched for Video-on-Demand (VoD) research methods. This network is created by Ramon Lobato, Karin van Es, and Jannick Sørensen. The site aims to support researchers investigating VoD platforms and methodologies. 🌐 Although still under development, it’s already up and running at vodmethods.net . The network was born from the Innovative Methods for Video-on-Demand Research event on 12 September 2024. The event, organized by Karin van Es from Utrecht University and Ramon Lobato from RMIT University, gathered VOD researchers to discuss the latest methodological innovations in VoD studies. Tim Raats and Catalina Iordache were invited to present their research on a methodological toolkit. 💡 Discover more about the SRP Media researchers here . One of the outstanding features of the new website is the Reading list . This resource is a living document, updated with essential readings on a variety of themes related to VoD research. It’s designed to serve both experienced researchers and students, providing a comprehensive starting point for anyone diving into the complexities of VoD platforms and their societal impact. As the network grows, so will the reading list, which we hope will become a go-to resource for the entire community. 📚 Next to the reading list, make sure to have a look at the Toolkit , and keep an eye on the site as it continues to develop. This is just the beginning for what promises to be an invaluable resource for VoD research. Previous Next New research network for VOD research methods

< Back Media policymaking and multistakeholder involvement: Matching audience, stakeholder and government expectations for public service media in Flanders Van den Bulck, H. & Raats, T. (2023). Media policymaking and multistakeholder involvement: Matching audience, stakeholder and government expectations for public service media in Flanders. European Journal of Communication , 38 (2), 132-147. https://doi.org/10.1177/02673231221112199 This contribution analyzes government, opposition, public service media, media stakeholders and audience views regarding the role and remit of public service media in the run-up to and their impact on the renewal of the 2021–2025 management contract between public broadcaster VRT and the Flemish Government. Results show that, despite a shifting media ecosystem and academics and government pushing for fundamental reform, audiences and most stakeholders’ views stick to a centralised, broad and multiplatform public media institution. Moreover, they expect public service media to solve ever more media and societal issues (e.g. fake news) within a shrinking budget. The case illustrates how mature, evidence-based multistakeholderism pushes public service media to meet an increasingly challenging set of expectations, hampers both public service media and government to build a well-balanced, long-term vision of public service media's role and, instead, pushes them to pursue their own agenda. Lastly, an evidence-based process also suffers from issues of validity of the data. Stakeholders, Platforms, Public Service Media Previous Read the article Next

< Back Platforms and Exposure Diversity: Towards a Framework to Assess Policies to Promote Exposure Diversity Ranaivoson, H. R., & Domazetovikj, N. (2023). Platforms and Exposure Diversity: Towards a Framework to Assess Policies to Promote Exposure Diversity. Media and Communication , 11 (2), 379-391. https://doi.org/10.17645/mac.v11i2.6401 The fragmentation of consumption and algorithms’ increasing impact on how content is recommended and displayed makes it even more important to analyse and promote exposure diversity, i.e., the extent to which audiences are exposed to, discover, and engage with diverse content. Although there is a growing literature addressing how to define media diversity in the context of the challenges posed by platformisation, this article translates the normative dimensions into a framework for operationalising exposure diversity into a tangible policy goal, taking into account datafication and its consequences in terms of increasing data requirements towards platforms. The main objective of this study is to analyse initiatives to assess exposure diversity in the platform era and to discuss how such assessment could be improved, particu‐ larly for policy initiatives. This involves addressing several challenges of existing approaches for the assessment of exposure diversity related to defining an appropriate frame of reference, determining the degree of diversity required, dealing with data transparency issues, and promoting user autonomy. To achieve this, we propose a framework for analysing initiatives aimed at assessing and promoting exposure to media diversity. Our framework is composed of four key features: measures (type of initiative), metrics (quantifying exposure diversity), data collection methods, and data requirements. We apply this framework to a set of 13 initiatives and find that policy initiatives can benefit from adopting metrics based on distances and experimenting with data collection methods. Platforms Previous Read the article Next

< Back Exploring Ethical and Regulatory Challenges of AI Integration in European Union Newsrooms Lefèvre, B., Errando, A., Afilipoaie, A., Ranaivoson, H., & Wiart, L. (2025). Exploring ethical and regulatory challenges of AI integration in European Union newsrooms. Media Studies , 16 (31), 31-55. Artificial Intelligence (AI) has significantly reshaped the news media landscape in the fast-evolving digital ecosystem. While recent legislative measures and industry self-regulation initiatives have emerged to address AI’s implications, the ethical challenges associated with its adoption in European Union (EU) newsrooms and the ways news media organisations navigate regulatory frameworks and internal initiatives remain underexplored. This study aims to bridge this gap by examining the conditions under which AI tools are integrated into media organisations from both organisational and regulatory perspectives. Drawing on an analysis of 30 key documents and 41 in-depth interviews with media professionals and regulatory experts across three EU markets, the study uncovers key findings: a limited discourse on AI use in media organisations, significant disparities in AI implementation practices, and an absence of clear, actionable regulatory guidelines for addressing considerations in AI-driven news content. Platforms Previous Read the article Next

< Back 23 Oct 2024 Sharing our reflections on the ACM RecSys Conference 2024 Some researchers of the SRP Media team recently attended the 18th ACM Conference on Recommender Systems (RecSys 2024) in Bari, Italy. Known for its stunning coastline and sunny weather, Bari provided the perfect backdrop for a gathering of both academic and industry experts in recommender systems. Beyond indulging in the delicious Italian cuisine and navigating the city’s chaotic traffic, our team traveled to Bari to share our latest research, gain insights from the inspiring talks, and engage with the Recsys community. SRP Media's Contributions at RecSys 2024 A previous blog post already summarizes all of the SRP Media team's contributions to the conference. Annelien and Lien each gave a learning session at the RecSys Summer School, which took place the week before the conference. Ulysse attended the Summer School and then stayed another week in beautiful Bari to present a demo poster at the conference. During the conference, Brett participated in the doctoral symposium, where he presented his initial findings and received valuable feedback from leading experts. In addition, there was also a poster for Hanne and Annelien's paper which got shortlisted for the Best Short Paper Award! The RecSys Summer School: Gaining Insights Although RecSys 2024 officially kicked off on October 14th, some of our team members arrived in Bari a week early to attend the ACM Europe School on Recommender Systems. On the program there was Lien Michiels 's lecture on Best Practices for Offline Evaluation and Annelien Smets ' course on Digital Economics and Recommender Systems . Their insights offered a refreshing perspective amidst the buzz around transformer-based algorithms. Ulysse, as a participant in the Summer School, had the chance to connect with fellow researchers from around the world, fostering an environment of collaboration and learning. The Main Conference: Workshops, Presentations, and Keynotes The RecSys conference consists of two parts: on Monday and Friday, a series of workshops and presentations took place at the Politecnico University of Bari. From Tuesday to Thursday, the main track continues with an extensive program of lectures, presentations and posters at the magnificent Teatro Petruzzelli. Some of our highlights: Ulysse's highlights For Ulysse, one of the standout moments was the CARS workshop led by Dr. Rishabh Mehrotra from Sourcegraph. Drawing on his experience at Spotify, Dr. Mehrotra presented a thorough analysis of eight real-world recommendation challenges, offering invaluable insights into the practical applications of recommendation systems. Ulysse also appreciated a keynote from Mounia Lalmas, also from Spotify, as well as events like the Women in RecSys session, the Normalize workshop, and Netflix’s “Raising a Recommender System”—although his duties as a student volunteer required him to follow these online. Brett's reflections Brett's experience at the doctoral symposium was the outstanding highlight of the week. Thanks to valuable feedback from experts such as Julia Neidhardt, Markus Zanger and Bart Knijnenburg, Brett gained new clarity on his PhD project, particularly his development of the "Serendiptionnaire" a tool designed to evaluate how recommender systems promote serendipity. Throughout the main track, Brett found himself both impressed and challenged by the industry presence. While captivated by the venue and delighted by the food, he noted a strong focus on technical issues, which sometimes overshadowed a more human-centered approach. Mark Riedl’s keynote resonated with this perspective, particularly the idea that recommender systems should enhance user agency. Workshops like RecSoGood, AltRec , and the POPROX project emphasizing the importance of considering the societal implications of research on recommender system. Hanne’s first RecSys Attending the RecSys Conference in person for the first time was an exciting and enriching experience. The conference brings together researchers and industry practitioners from a wide range of sectors, creating a vibrant and open-minded community. The empowerment of women in this traditionally male-dominated field, with the Women in RecSys breakfast and session is very valuable to facilitate connections. Her own contribution, a short paper titled “It’s (not) all about that CTR: A Multi-stakeholder Perspective on News Recommender Metrics” was shortlisted for the Best Short Paper award. This nomination, together with the encouraging feedback on the poster, is a testament to the importance of interdisciplinary approaches in recommender system research. The need for bridging the gap between academia and industry echoed throughout the week. Final Thoughts: A Rich Exchange of Ideas and Inspiration RecSys 2024 in Bari offered not only an incredible learning experience but also an opportunity to engage with great minds in the Recsys field. From insightful workshops and keynotes to helpful feedback and new connections, our team left the conference feeling motivated and inspired. As we look ahead, the key takeaway is clear: recommender systems must evolve beyond technical optimization. The importance of fostering user agency, promoting social sciences in the research on recommenders, and bridging the gap between academia and industry applications is more pressing than ever. Previous Next Looking back at RecSys 2024

< Back Advancing Media Recommendations through a User-Centric Understanding of Preferences, Experiences, and Agency Media recommender systems increasingly shape how users discover, select, and experience audiovisual content. However, user preferences are often inferred from observable behaviors, such as ratings, clicks, and viewing histories. These indicators may not fully capture the multidimensional, dynamic, and context-dependent nature of preference or users’ expectations regarding relevance, content quality, and autonomy. This can create tensions between what users value, what they consume, and what platforms recommend. This project takes a user-centric approach to examining how preferences are expressed, negotiated, and conditionally enacted in users’ interactions with media recommender systems. It investigates how preferences are conceptualized and operationalized, how users experience algorithmic recommendations, and how they navigate and respond to recommendations in everyday media use. Ultimately, the project seeks to inform more responsive and beneficial approaches to personalization that better account for users’ diverse and contextual preferences, experiences, and agency. Understanding user preferences in recommender systems: A systematic review with a user-centric perspective in the audiovisual domain 2026 Read Frustrations with Access to Preferred Content: User Experiences with Bilibili’s Recommender Systems Reflected on Weibo 2026 Read Researchers on this project Dongxiao Li PhD Researcher Consortium partners China Scholarship Council China Scholarship Council 1/1 Tags Recommenders, User studies Previous Next

< Back 22 May 2025 Hanne presented her research on stakeholder alignment for news recommender system development Today at WMEMC, Hanne presented her work on what happens behind the scenes of news organizations that are developing in-house News Recommender Systems (NRS). Spoiler alert: it’s no smooth ride. As newsrooms explore the possibilities of algorithmic curation, it turns out there's a lot more involved than just building smart tech. The real challenge? Getting all organizational stakeholders on the same page: journalists, software developers, managers, and business units each have their own goals, logics, and some concerns or doubts about (personalized) news recommendations. That’s where Hanne's research comes in. By setting foot in large-scale commercial news organizations (in Belgium, the Netherlands, and Scandinavia) and speaking to different stakeholders. Those interviews allow to explore the intricate (and let's be honest, sometimes messy) process of getting all organizational stakeholders to cooperate and negotiate NRS. The result? Balancing on a tightrope while juggling goals, demands, and objections. 🔹 Balancing Act #1 : Aligning Stakeholders Different stakeholder groups - newsroom, commercial units, and tech development teams - hold different perspectives on the opportunities and risks of (personalized) recommendations. Opposing views and expectations might lead to tensions between departments, and different product owners (POs) are assigned the role of mediators. These POs are ambidextrous actors with knowledge, experience, and professional backgrounds, crossing the traditional stakeholder boundaries, which enables them to create a shared understanding, translate jargon and build bridges between distinct stakeholder groups. A multi-disciplinary, cross-functional team of different POs, each representing another stakeholder group, sits around the negotiation table, and they are in charge of the decision-making around NRS. 🔹 Balancing Act #2 : Exploration vs. Exploitation Organizations constantly have to decide: do we explore new features and test new capabilities, or do we focus on optimizing our current technologies for ROI? With limited resources, news organizations can’t do both at full speed, simultaneously. Therefore, the team of POs needs to decide on which actions should be undertaken and, in doing so, develop a strategic roadmap navigating the tricky waters of budget and time allocations, and find that sweet spot between experimentation and exploitation. Both balancing acts are intertwined, as good alignment between stakeholders facilitates how these trade-offs are made. And the decisions made about the strategic roadmap for NRS development, in turn, influence whether stakeholders stay aligned or drift apart. It’s a dynamic, ongoing dance. The big takeaway? We need better, smarter ways to manage the decision-making on AI innovation in newsrooms, whether that relates to NRS or generative AI. This study is a first step in that direction — and just the beginning of figuring out how to align perspectives, reduce friction, and get these powerful tools off the ground in ways that actually work. Because building a scalable and effective recommender system is hard. But getting different stakeholders to agree on what the NRS design such look like? That’s where the real challenge lies. Previous Next Talking about Stakeholder Perspectives at WMEMC

< Back 24 Sept 2024 Our takeaways from the workshop part of the NWO-funded project "Rethinking news algorithms" On 23rd of September, Aina, Hanne, and Ulysse traveled to the Vrije Universiteit Amsterdam for the workshop "Best practices for responsible news recommender design," part of the NWO-funded project "Rethinking news algorithms". This one-day event aimed to foster collaboration between academia and industry, focusing on the ethical design of news recommender systems. Aina and Hanne presented initial insights from their systematic literature review that aims to uncover the organizational challenges, opportunities and social implications of news recommendation systems studied in Western media systems. One significant observation from the day highlighted the challenges in aligning all aspects of responsible news recommenders with diverse stakeholder goals. A key takeaway from our engaging conversations with colleagues and industry professionals is that diversity is frequently considered essential for the development of responsible recommenders, given the close relationship between the democratic function of news and media pluralism. The workshop enabled a cooperative sharing of ideas, fostering the co-creation of optimal strategies for responsible news algorithms. Attendees also pinpointed prospective research paths, underscoring the necessity of ongoing interdisciplinary collaboration to explore the societal implications of these systems. Previous Next Workshop: Best Practices for Responsible News Recommender Design

< Back Not that personal: A qualitative study of personalization practices on six European public streaming service Münter Lassen, J., Iordache, C., Martin, D., Bruun, H., Świtkowski, F., Nucci, A., & Johnson, C. (2026). Not that personal: A qualitative study of personalization practices on six European public streaming service. Television & New Media , 1–19. https://doi.org/10.1177/15274764251415356 This article presents key findings from a comparative empirical study of content recommendation on the audiovisual streaming services of six European public service media (PSM) organizations. Due to growing competition with international subscription streaming services, the editorial practices of PSM are developing, combining editorial curation with algorithmic recommendations. Contributing to the fields of PSM, platformization and scheduling studies, the article presents findings from a qualitative reverse engineering approach to interface analysis. We document how the majority of the services exercise what we characterize as light personalization, and we highlight explanations for why the degree of personalization on the PSM streaming services, in most cases, is still limited. The comparative approach enables us to identify the factors most commonly at play across diverse national contexts. The article adds to theoretical debates about how personalization is constructed in a PSM context and discusses whether algorithmic personalization conflicts with PSM values. Platforms, Recommender Systems Previous Read the article Next

< Back 28 May 2024 Help us by providing your music listening data Your playlists have the power to drive change. By donating your favourite tunes to the Fair MusE project, you're not just sharing good vibes – you're supporting artists who deserve fair recognition and compensation! SMIT , the research group at imec & VUB, is part of a wider interdisciplinary team of academic and industry partners whose mission is to raise awareness about how music algorithms, data collection, and exploitation models of social media and streaming platforms influence music creators and audiences. In a nutshell, the three-year Horizon EU funded project aims to promote fairness for a more sustainable and competitive music ecosystem in Europe. If you are a user of Spotify, Apple Music or YouTube Music, be part of the change and donate your music streaming data here ! Let's build a more inclusive and equitable music industry, one playlist at a time! For more information you can reach out to info@fairmuse.eu . For updates stay tuned on LinkedIn . Previous Next Calling all music lovers!

< Back 8 Oct 2024 Our contributions to the 18th ACM Conference on Recommender Systems From 14 until 18 October 2024, the 18th ACM Conference on Recommender Systems takes place in Bari, Italy. Our SRP Media researchers have various contributions to this conference, as well as the preceding ACM Europe School on Recommender Systems . ACM Europe School on Recommender Systems 2024 During the ACM Europe School on Recommender Systems, leaders in the field give lectures on the practice, research, and state of the art in Recommender Systems. This year, there will be 120 attendees among which 90 students. Click here for the full program . On Tuesday 8 Octobe r, Lien Michiels will teach a lecture on Best Practices for Offline Evaluation . On Friday 11 October, Annelien Smets lectures a course on Digital Economics and Recommender Systems . The ACM Europe School on Recommender Syste ms is organized by Ludovico Boratto (University of Cagliari, Italy) , Cataldo Musto (University of Bari, Italy) , and Fedelucio Narducci (Polytechnic University of Bari, Italy) . How to Evaluate Serendipity in Recommender Systems: the Need for a Serendiptionnaire During the Doctoral Symposium , Brett Binst will present his work from the Serendipity Engine project. Recommender systems can assist in various user tasks and serve diverse values, including exploring the item space. Serendipity has recently received considerable attention, often seen as a way to broaden users’ tastes and counteract filter bubbles. However, the field of research on serendipity is fragmented regarding its evaluation methods, which impedes the progress of knowledge accumulation. This research plan proposes two studies to address these issues. First, a systematic literature review will be conducted to provide insights into how serendipity is currently studied in the field. This review will serve as a reference for novice researchers and help mitigate fragmentation by presenting a thorough overview of the field. This systematic literature review has already revealed a significant gap: the lack of a validated, widely accepted method for evaluating serendipity. Therefore, the second part of this research plan is to develop a validated questionnaire, the serendiptionnaire, to measure serendipity. This tool will provide a ground truth for evaluating serendipity, aiding in answering fundamental questions within the field and validating offline metrics. GenUI(ne) CRS: UI Elements and Retrieval-Augmented Generation in Conversational Recommender Systems with LLMs [demo] During the poster session on Tuesday, Ulysse Maes will present joint work with Lien Michiels and Annelien Smets . Previous research has used Large Language Models (LLMs) to develop personalized Conversational Recommender Systems (CRS) with text-based user interfaces (UIs). However, the potential of LLMs to generate interactive graphical elements that enhance user experience remains largely unexplored. To address this gap, we introduce “GenUI(ne) CRS,” a novel framework designed to leverage LLMs for adaptive and interactive UIs. Our framework supports domain-specific graphical elements such as buttons and cards, in addition to text-based inputs. It also addresses the common LLM issue of outdated knowledge, known as the “knowledge cut-off,” by implementing Retrieval-Augmented Generation (RAG). To illustrate its potential, we developed a prototype movie CRS. This work demonstrates the feasibility of LLM-powered interactive UIs and paves the way for future CRS research, including user experience validation, transparent explanations, and addressing LLM biases. It’s (not) all about that CTR: A Multi-Stakeholder Perspective on News Recommender Metrics During the poster session on Thursday, Hanne Vandenbroucke will present joint work with Annelien Smets . Recommender systems are increasingly used by news media organizations. Existing literature examines various aspects of news recommender systems (NRS) from a computational, user-centric, or normative perspective. Yet research advocates studying the complexities of real-world applications around NRS. Recently, a multi-stakeholder approach to NRS has been adopted, allowing to understand different stakeholder perspectives on NRS development and evaluation within the news organization. However, little research has been done on the different key performance indicators (KPIs) and metrics considered valuable by different stakeholders. Based on 11 interviews with professionals from two commercial news publishers, this paper demonstrates that stakeholders prioritize distinct KPIs and metrics related to the reach-engagement-conversion-retention funnel. The evaluation of NRS performance is often limited to short-term metrics like CTR, overlooking the multiplicity of stakeholders involved. Our findings reveal how different purposes, KPIs, and metrics are valued from the journalistic, commercial, and tech logic. In doing so, this paper contributes to the multi-stakeholder approach to NRS, advancing our understanding of the real-world complexity of NRS development and evaluation. Navigating Discoverability in the Digital Era: a Theoretical Framework During the MuRS: Music Recommender Systems Workshop 2024 , Rebecca Salganik will present joint work with Valdy Wiratama , Heritiana Ranaivoson and Adelaida Afilipoaie (FairMuse). The proliferation of digital technologies in the distribution of digital content has prompted concerns about the effects on cultural diversity in the digital era. The concept of discoverability has been presented as a theoretical tool through which to consider the likelihood that content will be interacted with. The multifaceted nature of this broad theme has been explored through a variety of domains that explore the ripple effects of platformization, each with its own unique lexicography. However, there is yet to be a unified framework through which to consider the complex pathways of discovery. In this work we present the discovery ecosystem, consisting of six individual, interconnected components, that encompass the pathway of discovery from start to finish. Workshop on Normative Design and Evaluation of Recommender Systems (NORMalize) On October 18, Lien Michiels is co-organizing t he Second Workshop on Normative Design and Evaluation of Recommender Systems. Previous Next The RecSys Conference 2024

< Back 31 Mar 2025 Ulysse attended the conference and presented at the AXAI workshop Pristine beaches, a bright turquoise sea, impressive rock formations and pink flamingo’s… Cagliari has it all. Despite the undeniable beauty of Sardinia’s capital, the most memorable element of my stay was however related to Intelligent User Interfaces. You read that right: I was lucky enough to attend the yearly conference on Intelligent User Interfaces (IUI) - which turned out to be an exceptionally inspirational and instructive experience. It shouldn’t come as a surprise that my interest was sparked by the central theme - the conjunction between Artificial Intelligence (AI), and how User Interfaces (UI) can optimally accommodate for them. But the real experience went above and beyond my initial expectations. In this article, I break it down in 3 key elements. 1. The overarching theme: AI should benefit the user Believe it or not, but the word “user” was mentioned even more often than “LLM” (see paragraph 3; “There is no escaping LLMs”). Shouldn’t be too surprising for a conference focusing on user interfaces, right? Well, it struck me that this community really tries to model user preferences and -needs accurately, for which it draws a lot more often from cognitive and behavioral research than is common in computer science. As I’m also considering to conduct user experiments to evaluate the effects of explanations on the complex experience of the system, this focus meant that I learned a lot about suitable methodologies. During the course of the week, my belief in the importance of putting the user central has been fortified as well. Now more than ever, it is important to guard user agency and autonomy, and to prevent over-reliance on increasingly capable and connected AI systems. The current generative AI boom raises a lot of concerns among various affected stakeholders. Sometimes, it feels like these tools are at least partly exploitative in nature - a concern that was vividly expressed by professor Giulio Jacucci in his opening keynote at the HAI-GEN workshop. 2. The AXAI workshop: Explanations should be adaptive to be impactful Throughout the presentations in the AXAI (Adaptive eXplainable AI) workshop session, I learned that adaptivity can be geared towards multiple users and use cases, and is therefore often interpreted differently by various researchers and fields. Some focus on adapting explanations to the user at hand, in terms of content, or complexity. Others consider adaption to context. In any case, it seems that LLMs offer a great avenue to increase the level of adaptation in many cases, although this may spark concerns regarding trustworthiness. Related to that topic, I presented our paper ( link ), co-authored with Lien and supervised by Annelien. What set it apart was mainly its ridiculously long title: “Mitigating Misleadingness in LLM-Generated Natural Language Explanations for Recommender Systems: Ensuring Broad Truthfulness Through Factuality and Faithfulness”. Not exactly the title of Sabrina Carpenter’s next hit, but it certainly sparked some lively discussion. I opened my presentation with a straightforward example of gender bias in a job recommender: it generated the exact same “high-quality” explanation for two different users—a female designer recommended design jobs and a male designer steered toward management roles. This example served as a way to illustrate the benefit of incorporating uncertainty and interactive counterfactuals into explanations, to enable greater transparency and scrutability. I recognized a similar call for communicating uncertainty as a means to obtain transparency in Prof. Q. Vera Liao’s keynote on Thursday. I was also pleased to see Prof. Turchi present “Talking Back - human input and explanations to interactive AI systems“, an inspiring study on interactive counterfactuals using SHAP values as sliders. It sparked my interest in further exploring the integration of interactive explanations with SHAP in content-based recommender systems. My presentation then proceeded with stating the difference between explanations and justifications, and how LLMs enable the generation of plausible justifications at scale. This can pose a problem of unfounded trust in recommender systems, especially since earlier research has shown that the mere presence of explanations already enhances the trust and item acceptance. This underscores the importance of truthful explanations - which brought me to the core of my presentation: to assess “truthfulness”, we should agree on a definition and an operationalization. This is where I propose to frame truthfulness as “providing accurate information”, consisting of both factuality and faithfulness. Unfortunately, we often see that different research disciplines consider different aspects of truthfulness: while computer science works often focus on factuality, social science merely investigates faithfulness. Luckily, as Krzysztof Gajos mentioned in Wednesday’s morning panel, the field of IUI is well-positioned to bridge these different perspectives into truly useful interfaces. Slide from AXAI presentation: "Defining truthfulness" I proceeded with mentioning the 4 evaluation perspectives to assess explanation quality in recommender systems, as proposed by Ge et al., while also mentioning the 7 explanation goals as defined by Tintarev in 2007. I was curious if the audience would deem this evaluation method useful outside of recommender systems as well, and it indeed proved to be fertile ground for discussion. While most workshop participants agreed that the dimensions could be useful, the explanation goal of “persuasion” was contested, as this should be seen as a side-effect or external result, not something to optimize for. To end my presentation, I went over some possible mitigation strategies that focus on prompting, interface and model-based approaches (all are further discussed in the paper). Slide from AXAI presentation: "Mitigation strategies for LLM-generated misleading explanations" Besides the "Talking back" paper, the presentations most relevant to my area of interest in this session were "Toward a Human-Centered Metric for Evaluating Trust in Artificial Intelligence Systems" and "'Loss in Value': What it reveals about WHO an explanation serves well and WHEN". By the way, in the morning, I attended another great workshop: HAI-GEN, in which the shift to an intent-based paradigm was stressed multiple times. I was very happy to hear this, as I’m also convinced that we are effectively moving to a new way of interaction with our systems, were we move from command-based interaction towards a more natural way of communicating our goals to digital systems. As far as I’m aware, the term “intent-based interaction” was coined by Jakob Nielsen in a now-famous blog post . I find this very inspiring and consider it one of the main guiding threads for my research. There is no escaping LLMs As announced in the opening talk, I indeed noticed that LLMs were ubiquitous. From self-improving LLM-agents that learn to play Minecraft, to LLMs that optimize meeting schedules: a lot of authors (including myself) reported on the promises and perils of incorporating language models to enhance interactivity, accessibility or automation. The popularity of LLMs should however not be mistaken for devotion. Many times I realized, as a scientific community, we have the privilege to be critical, and look beyond the hype - so that we can report on both the opportunities as well as the limitations of LLMs. That doesn’t mean that we shouldn’t be excited about the seemingly endless possibilities of this new technology of course. But as was rightfully highlighted by Prof. Burnett, we should consider it in a responsible way. By doing so, we can provide valuable insights that help to steer the development and implementation of AI-models beyond purely profit-driven goals. Conclusion: An inspiring conference at the heart of my research interests This post could have been much longer if I had included all my notes. I'll end it here for now, but in the coming weeks, I hope to frequently revisit these notes to reignite the inspiration I felt from attending the conference, listening to speakers, admiring impressive research projects, and meeting some of the most influential HCI researchers. Two memorable encounters will forever remain in my mind: the first was a lunch conversation with Professor Ted Selker, the creator of the well-known red pointing stick on my ThinkPad (though there's a more familiar term for it; if you know, you know). The second? A close encounter with a startled yet adorably cute beaver during a run around Cagliari’s stunning salt mines. I hope to reconnect with many of the inspiring individuals I met at the conference in the future. In the meantime, let’s continue our ongoing pursuit of the perfect intelligent interface. Previous Next Looking back at IUI 2025