Beyond clicks! Value-driven recommendations
This project is a collaboration with DPG Media, the AI Labs at Utrecht University (UU), and Paris Lodron University Salzburg (PLUS)—with Karin van Es and me as Co-PIs.
Abstract
Many news platforms provide users with recommendations for a personalized experience that fits their interests and needs. Providing users with relevant news and a good experience are key to such platforms. However, aside from the consumption of news articles and user loyalty, additional values are relevant; for instance, (political) neutrality, factual news, understandable news, content diversity, and the reach of a broad audience.
Within this research project, we plan to (i) identify relevant values for both the platform provider and the users; (ii) translate and operationalize a selection of these relevant values into new metrics; (iii) test and evaluate recommender systems considering the identified values and metrics.