News and media recommender systems by Ozlem Ozgobek and Daan Odijk #rsss2023 Research Centre for Responsible Media Technology

Day 4 - Session 11 of 2023 ACM RecSys Summer School: Media recommendation by Prof. Özlem Özgöbek and Daan Odijk

Excited that RecFusion is out arXiv 🎉 We propose a Bernoulli Diffusion model formulation for binomial data in 1D or 2D, applied to recommendation. paper: arxiv.org/abs/2306.08947 code: github.com/gabriben/recfu… co-authors: Olivier Jeunen Samuele Papa Samarth Bhargav Daan Odijk Maarten de Rijke

We are happy to welcome another PhD student in our lab: Maya Nachese. She will work at University of Amsterdam on automatic TV subtitling.

Our recent paper "VideolandGPT: A User Study on a Conversational Recommender System" was published at KARS2023 (5th Knowledge-aware and Conversational Recommender Systems Workshop with #RecSys2023, by Mateo Gutierrez Granada, Dina Zilbershtein, Daan Odijk and Francesco Barile

Is time to introduce one new member of our lab: Daniella Pauly Jensen, who joined us in mid-August, as TAIM PhD candidate based at UM's Faculty of Arts & Social Sciences (FASoS).

Our Mirela Popa is finally complete and we had our first PhDs kickoff meeting on 28th Sept. at RTL. We are ready to embark on the fascinating trip of developing video summaries, fair ad recommendations, personalization, automatic subtitling, while ensuring fairness and diversity!

❗Starting soon! Our panel discussion on 'Media Content on Responsible AI Training', co-organised with Open Future Foundation, live at the PublicSpaces conference. Livestream: conference.publicspaces.net/live/7

Then, on Friday, Dina will be presenting her paper with Francesco Barile, Daan Odijk, and @navatintarev: "Bridging the Transparency Gap: Exploring Multi-Stakeholder Preferences for Targeted Advertisement Explanations" during the IntRS workshop! In this paper, they conducted a user (3/7)