Zoekresultaten

< Back From Experimentation to Strategy: The Ambidextrous Balancing Act of Developing News Recommender Systems Vandenbroucke, H., & Smets, A. (2025). From experimentation to strategy: the ambidextrous balancing act of developing news recommender systems. Journal of Media Business Studies , 1–28. https://doi.org/10.1080/16522354.2025.2590880 Despite advances in news recommender systems (NRS) research, their application in news organisations remains limited, hampered by practical obstacles and organisational challenges. This study applies the multi-stakeholder framework alongside ambidexterity theory to analyse the complex decision-making dynamics in applying NRS within two large commercial news organisations. Based on 11 in-depth interviews with different stakeholders, two “balancing acts” emerged. The first is stakeholder alignment through two strategies: knowledge management with dummy-proof presentations and “showcasing success” to help increase stakeholder buy-in for NRS. The second balancing act is managing trade-offs between exploration and exploitation in a roadmap developed by the multidisciplinary team of product owners (POs), ensuring a balanced allocation of resources for NRS innovation and optimisation. This research shows how organisational ambidexterity supports successful technology integration and provides insights for implementing emerging technologies in complex multi-stakeholder news media environments. Recommender Systems, Stakeholders, Newsmedia Previous Read the article Next

< Back Wrap-up of the RecSys Summer School 16 Jun 2023 Our key takeaways from the Recommender Systems Summer School in Copenhagen From 12 to 16 June, several researchers from the Media Economics & Policy Unit took part in the Recommender Systems Summer School in Copenhagen. During this Summer School, academics and industry leaders lectured on the practice, research, and state of the art in recommender systems. The lectures covered a broad range of topics from an algorithmic as well as a methodological perspective, including hands-on sessions. This week brought many interesting insights for our strategic research program on recommender systems and this short report highlights some of the key takeaways. Not an algorithm, but a system Recommender systems encompass more than just the algorithm itself. They require the thoughtful implementation of (a combination of several) algorithms that align with specific domain objectives, account for optimal user experience in different contexts, and seek to create mutual value for various stakeholders It is crucial to recognize that there is no one-size-fits-all approach to recommender systems. The strategy employed must be tailored to the specific domain in which the system operates, considering the unique objectives, economic factors, values and challenges inherent to that domain. For instance, in the realm of news, evaluating the system's quality necessitates considerations such as diversity and serendipity, speed and coverage. The recommender system design must consider the overall user experience. This involves understanding the various features and styles influencing users’ interaction with the platform, as well as analyzing the user journey and recognizing the impact of contextual dependencies on their preferences and needs. In both academic literature and practical applications, the primary goal of recommendation systems is to create value. This entails a reciprocal relationship between user values and business values. By addressing user needs and providing quality personalized recommendations that go beyond mere accuracy the system increases short-term engagement. Furthermore, these systems aim to cultivate long-term loyalty and build strong relationships with the target audience. In summary, well-designed recommendation systems enhance user engagement by offering personalized recommendations, optimizing the user experience, reducing information overload, fostering serendipitous discovery, continuously learning and improving, integrating social features, and thereby generate added value for business through increased user satisfaction, retention, and potential revenue growth. End of the artificial clean cut between content? Media mergers are changing the industry. One of the many examples is how RTL XL will become a part of Videoland. The question rises how the video-on demand platforms can blend different content types such as movies, series, TV programs, short clips and livestreams in an appealing way. An answer to that question could be answered by formulating the optimal user experience through recommender systems. Currently, the company is analyzing user behaviour in order to develop models that can be used in online user experiments. A next step will be to start A-B testing to create the optimal recommendation model. The goal of VOD platforms is to have loyal visitors, but it is a metric that moves very slowly. Currently, the recommendation system of the RTL is built up on three different types of recommendations: content-based, popular within the genre and collaborative filtering. Interesting fact: Personalized swimming lanes compared to editorials swimming lanes generate 30 min more viewing time per active user per month. The main goal of “this recommender optimizing project” is to work towards continuous loops. Sequential recommender systems are different in that sense that they convert user’s behavior trajectory into recommended items or services. It takes into account the current and recent preferences of a user for a more accurate recommendations. Implementing this new form of recommender system will be one of the key factors to generate a user interface with “blended content” that answers the customer’s needs. A critical stance in the evaluation of recommenders Being grounded in dominantly quantitative forms of assessment, recommender system evaluation needs to pay sufficient attention to real-world significance of numerical results and to whether outcomes actually make sense in applied cases. For instance, is an overall increase in prediction accuracy of a recommendation an accomplishment when large parts of the user base still receive bad recommendations and their preferences remain ill-defined? A call for qualitative sense-making of quantitative evaluation outcomes was certainly made at the Summer School. Also, research papers that seek to evaluate recommender systems often focus more on (incremental) increases in performance percentages than on a solid basis for their actual evaluation. In terms of relevance for the academic field, consequently, little contributions are made. Papers that are characterised by vagueness and technical complexity thus prevent real progress and cannot form the basis for further research. The lack of valuable longitudinal research in the domain of recommender systems can also be related to this. Lastly, by acknowledging that recommender systems impact not only the intended end-user and that their complex nature implies influences also on those not directly involved, we argue that multiple-stakeholder considerations should be the norm. We realize that the inherent complexity of recommender systems makes this a difficult endeavour. But in our attempts to evade the McNamara fallacy and to research in the most holistic way possible, the aim should always be to keep a multi-stakeholder involvement top-of-mind. Not solely in phases of design or evaluation, but as a constant reflective thought from the outset and throughout. Previous Next

About SRP Media SRP Media represents the third Strategic Research Programme on media economics at imec-SMIT, Vrije Universiteit Brussel . Its focus lies on understanding algorithm-driven media industries and how they reshape value in small markets. SRP Media c lusters four PhD trajectories and affiliated research projects that deal with Streaming, Recommenders and Platformization in European Media Industries. By doing so, SRP Media aims to foster collaborations and knowledge exchanges across several national and international research projects, and to function as one centralised hub towards external stakeholders. Research projects Streaming, Recommenders and Platformization in European Media Industries The research is conducted within the broader field of media and communication studies, more particularly in the research fields of media economics, political economy, innovation studies, and media policy. The importance of the research is confirmed by existing academic research, current media company practices, and ongoing policy developments in the field. Recent academic research has already been exploring the topic of algorithms in media industries. However, this body of work remains limited, mainly due to the ongoing changes in algorithm use, as well as the limited data publicly available on how algorithms and content recommendations are made in practice. In order to contribute to scientific evidence on the topic, as well as to industry- and policy-related debates and developments, the SRP follows three main research industries, namely news media, public service media, and global and domestic streaming platforms. Each industry will be analysed from four perspectives: media companies, media content, media audiences, and media policy. It also aims to analyse the interplay between the three, to identify their effects on audience consumption, and the ways in which algorithms are used and regulated. The research combines qualitative and quantitative methodology and specifically focuses on small media markets. It also applies case studies, both individually and in comparative analyses, either between different media service providers, or between different EU Member States. The programme will provide media stakeholders with the necessary knowledge on the usage and effects of algorithms on media production, distribution, and consumption. The findings will also help policymakers to formulate legislation that will accurately measure and efficiently regulate the implementation and use of algorithms and recommender systems. Principal Investigators Pieter Ballon Supervisor Tim Raats Supervisor / Track Lead Annelien Smets Track Lead Wendy Van den Broeck Track Lead Meet our team You might also be interested in ... ... fellow knowledge hubs at imec-SMIT, Vrije Universiteit Brussel related to media and technology. Nieuwsgebruik.be Mediapunt Mediawijs Knowledge Center Data & Society

< Back Annelien Smets Research Professor Annelien.Smets@vub.be Annelien is a senior researcher at SMIT in the Media Economics and Policy unit. Her research centers around personalization and recommender systems, and their value in media markets. Annelien holds a PhD in Media and Communication Studies (VUB) on the topic of serendipity in recommender systems and smart cities. She holds a master’s degree in Information Management (2016) and Artificial Intelligence (2017) from KU Leuven. Currently, Annelien is Research Professor at Vrije Universiteit Brussel, where she teaches courses on media economics, digital economics, digital business models, and digital innovation management. She is also co-chair of the Serendipity Academic Researchers Network, part of the Serendipity Society. Visit my research profile

< Back Our work-in-progress at DBWRS 2023 18 Dec 2023 Explore our work-in-progress presentations at DBWRS2023 Last week our team attended the first edition of the Dutch-Belgian Workshop on Recommender Systems. DBWRS 2023 proved to be a dynamic event for the exchange of ideas, insights, and ongoing research. Among the highlights were the six compelling work-in-progress posters presented by our researchers. In this blog post, we are excited to share a glimpse into these projects, providing a brief overview of the groundbreaking work our team is currently undertaking. 1. What Will We Be Streaming Tonight? And Why? This project delves into the fascinating world of streaming preferences. The Living Lab project explores the affordances that influence users' choices in content consumption. From binge-worthy series to thought-provoking documentaries, we're unraveling the intricate tapestry of streaming behaviors. 2. Newsroom Realities: An Exploration of Changing Dynamics in News Organizations in Relation to Recommender Systems In a rapidly evolving media landscape, Hanne's PhD-journey investigates the impact of implementing recommender systems on newsroom dynamics. How do different stakeholders wihtin news organizations look at the potential opportunities and risks of recommendations and personalization? 3. Assessing the Potential of Large Language Models for Personalized Explainable Recommendations in Media Large language models have revolutionized natural language processing. In his PhD, Ulysse will explore their potential for personalized and explainable recommendations in the media domain. Uncover the methodologies and insights that pave the way for a more transparent and user-centric recommender system. 4. Intention and Behavior: A Systematic Review of Literature on Users Preferences in Recommendation Systems Understanding users is at the heart of designing effective recommendation systems. Through an systematic literature review, Dongxiao sheds light on the intricate interplay between user intentions and actual behavior. 5. Discovering the Rhythm: The Impact of Online Platform Recommender Systems on Music Discoverability Music is a universal language, and our researchers in the FairMuse project are exploring how online platform recommender systems influence the discoverability of music. From algorithmic playlists to tailored suggestions, we're uncovering the rhythm that shapes users' musical exploration. 6. Gatekeeping in the Digital Age: Newsroom Resistance to News Personalization As news personalization becomes more prevalent, Aina investigates the resistance to recommender systems in newsrooms in Spain as a part of the Algepi-project . These work-in-progress posters showcase the diversity and depth of ongoing research of the SRP Media team. Together, we navigate the ever-changing landscape of recommender systems, digital platforms and streaming services. Previous Next

< Back Second Annual ALGEPI Workshop 19 May 2025 Insights on AI, Innovation and Media Regulation On April 23rd, 2025, the Université de Namur hosted the second ALGEPI Annual Workshop , bringing together researchers, students, and industry experts to discuss how AI is shaping the media landscape. The day was packed with presentations on artificial intelligence, algorithmic recommender systems, media regulation, and user-centric innovation. Some highlights: Prof. Heritiana Ranaivoson kicked off the day by outlining the ALGEPI project, which is a collaborative effort aiming to tackle the challenges AI poses to epistemic welfare. Dr. Lien Michiels dissected the elusive concept of filter bubbles, stressing the need for diverse, but standardised research methods and validated metrics that consider all stakeholders to gather robust evidence and make meaningful normative judgements about diversity and recommender systems. Aina Errando , Michelle Kulig, and Hanne Vandenbroucke shared some findings from their comparative systematic literature review on the multi-stakeholder challenges and opportunities of news recommender systems in newsrooms. Their work highlights the need for more interdisciplinary and multi-method research to address both organisational and societal challenges, and also to explore the opportunities of news recommender systems more holistically. For the full report on the 2nd ALGEPI workshop, have a look at https://www.algepi.com/ai-innovation-media-regulation-insights-ii-annual-workshop/ Previous Next

< Back Dubbing wars: localisation strategies of transnational streaming services for Spanish ‘original’ works Gallo, P., & Iordache, C. (2026). Dubbing wars: localisation strategies of transnational streaming services for Spanish ‘original’ works. Convergence: The International Journal of Research into New Media Technologies , 1-17. https://doi.org/10.1177/1354856526141627 This research investigates the intricate process of localising subscription video-on-demand (SVOD) services on a transnational scale, with a specific focus on the translation of Spanish ‘original’ and ‘exclusive’ works of Netflix and Prime Video. As SVOD companies expand their services internationally, the necessity of adapting to diverse national markets becomes paramount. Language emerges as a pivotal factor in this process: audiovisual works must provide adequate subtitling and dubbing options to address the diverse linguistic nuances of international audiences. The research aims to analyse the language localisation strategies for Spanish works in transnational SVOD services, by examining the dubbing and subtitling options offered. The article provides a mapping of the strategies and partners adopted by these companies focussing on a sample of local works. Additionally, by collecting data from the two SVOD services and through interviews with dubbing studios, the research analyses mergers, acquisitions and power relations between dubbing companies, as well as the impact on competition generated by transnational streaming services. Both Netflix and Amazon have increasingly added more language options for content offerings as part of their internationalisation, which has favoured the proliferation of a specific type of company and business model in the dubbing sector. This consists of transnational corporations that have provided localisation services in the wake of the internationalisation of streamers, and that have managed to keep up with the demands through mergers and acquisitions with studios in foreign countries. Platforms Previous Read the article Next

< Back Lien Michiels Postdoctoral Researcher lien.michiels@vub.be Lien Michiels is a Postdoctoral Researcher at imec-SMIT, Vrije Universiteit Brussel and the Adrem Data Lab in the Department of Computer Science at the University of Antwerp. Her research focuses on diversity, discovery and filter bubbles in recommender systems. Previously, she combined her PhD research with her work as a Machine Learning Engineer at Froomle , who provide their recommendation platform as a service to media companies around the globe. She obtained a master's degree in Mathematical Engineering at the KULeuven in 2017. Visit my research profile

< Back Our presentations at ECREA 1 Oct 2024 The SRP team presented their work at the ECREA 2024 Conference The 10th edition of the European Communication Research and Education Association Conference (ECREA) took place on 24-27 September 2024 in Ljubljana, Slovenia. The VUB’s Communication Sciences department had a large delegation of over 25 researchers at the conference, including several SRP team members. On Wednesday 25 September, Tim Raats presented a paper entitled ‘ Ensuring visibility of European public service media. An analysis of policy, industry and academic views on prominence measures in Belgium ’, as part of the session ‘Public service and the public interest in European Media’, in the Communication, Law and Policy section. The presentation was based on research conducted together with SRP colleagues Adelaida Afilipoaie and Pieter Van der Elst . On Thursday 26 September, Catalina Iordache presented research conducted together with Catherine Johnson (University of Leeds) on ‘ Balancing the scales between public service algorithms and editorial curation , the cases of BBC and VRT. The presentation was part of pre-constituted panel on ‘Comparative approaches to public service media disruption: The transformation of values, norms and prevailing structures in the age of platforms’, hosted by the Public Service Media in the Age of Platforms ( PSM-AP ) project. And on the final day of the conference, Friday 27 September, Noëmie Forest presented the paper ‘What will we be streaming tonight? And why? An analysis of motivations for VOD consumption in Belgium’, co-authored with Isabelle Puskas, Wendy Van den Broeck and Tim Raats. The presentation was part of the session on Audiovisualities and Audiences, in the section Audience and Reception Studies. Tim Raats and Catalina Iordache also presented their ongoing work on assessing discoverability and prominence on VOD services as part of the roundtable ‘ Re-conceptualising the television 'text' for the platform age: textual analysis, texts and interfaces ’, organised in the Television Studies section. As part of the business meetings, Catalina Iordache and Tim Raats have also be re-elected as vice-chairs of their respective sessions Media Industries and Cultural Production and Communication, Law and Policy . Previous Next

< Back 10 trends in streaming market 1 Mar 2024 Report of the Living Lab project Discover the first deliverable of the Streaming Affordances for Small Media Markets living Lab project, in which the research team outlines 10 key trends in the international streaming market that have an impact on developments in the Flemish market (the report is in Dutch). The 10 trends highlight a diversification of genre and payment modules, and a quest of streamers for revenue and profit rather than market dominance and number of subscribers. Trendrapport voor 'proeftuin Streaming Affordances for Small Media Markets .pdf Download PDF • 3.74MB Previous Next

< Back GenUI(ne) CRS: UI Elements and Retrieval-Augmented Generation in Conversational Recommender Systems with LLMs Maes, U. , Michiels, L. & Smets, A. , (8 Oct 2024). GenUI(ne) CRS: UI Elements and Retrieval-Augmented Generation in Conversational Recommender Systems with LLMs. Proceedings of the 18th ACM Conference on Recommender Systems. Bari: ACM , p. 1177-1179 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. Recommender Systems Previous Read the article Next

< Back Brett Binst PhD Researcher brett.binst@vub.be Brett Binst currently works on the ‘Serendipity Engine’ project as a PhD student. In this project, he studies serendipity in urban recommender systems. More specifically, he studies why system providers would design for serendipity, how they can design for it (through an affordances perspective) and the experience of serendipity in users of these urban recommender systems. Brett Binst acquired a Bachelor of science in Psychology at the VUB in 2020. Next, he completed his Master’s degree in Sociology in 2022. His masterthesis was a study into the preconditions and inequality of having an opinion about AI, inspired on Bourdieu’s theoretical framework laid out in La Distinction. Before starting on the Serendipity Engine project, he worked on the Barometer project, on behalf of the Koning Boudewijn Foundation in which he mainly performed quantitative analyses, comparing the labour market situation of migrants and natives. Visit my research profile