MT Designing an Interactive Recommender System to Test Users’ Strategies for Influencing Recommendations

In the contemporary information age, recommender systems (RSs) play a critical role in influencing online behaviour: from social media to e-commerce, from music streaming to news aggregators, individuals are constantly targeted by personalized recommendations suggesting contents that may interest them. Despite such diffusion, the extent to which recommendations influence users’ decisions is still underexplored, given that independent audits on the functioning of RSs deployed on online platforms are usually prevented by proprietary constraints. The Digital Services Act of the European Union [1] (DSA) is the first supranational regulation that sets specific transparency and auditing requirements for RSs used by online platforms with the aim of enhancing users’ self-determination: in addition to requiring platforms to explain the functioning of their RSs in the terms and conditions, it allows users to modify the parameters on which recommendations rely so to let them choose autonomously which kind of content they want to see.

The aim of this thesis project is to understand whether and how users would implement their new right to influence automated recommendations. Possible research questions can be: 
•    If users could modify the parameters of a RS, which features of the recommendation process would they find useful to intervene on?
•    To which extent the introduction of explanations of RSs as a standard, pop-up feature on the platform’s interface can enhance users’ awareness and competence in modifying the parameters of recommendations?
•    How can effective recommendations be designed without user profiling?

To answer such research questions, the student should build an interface-based simulation of RS (also implementing open-source models inspired from real-world algorithms used by online platforms) aimed at carrying user studies. The content of the recommendations can be proposed by the student among areas like e-commerce, social media, travels, music, news, as long as it falls in one of the domains of the very large online platforms (VLOPs) identified by the European Commission [2]. The interface should allow users to experiment with different settings of the parameters and features of RSs regarding type (collaborative, content-based, feature-based, non-personalized), policy (exploitative, explorative) and explanations of the recommendations. The student should have sufficient programming skills, acquaintance with interface design concepts and interest in the ethical and social implications of AI technologies.

This topic for a Master's Thesis will be co-advised by Wolfgang Wörndl and Matteo Fabbri (IMT School for Advanced Studies Lucca). Please send your application (brief CV, transcript of records and short motivation statement) to Wolfgang Wörndl (woerndl@in.tum.de).

[1] REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on a Single Market For Digital Services (Digital Services Act) and amending Directive 2000/31/EC: https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52020PC0825&from=en

[2] The full list can be found here: https://digital-strategy.ec.europa.eu/en/policies/dsa-vlops