MT Investigating User Control to Influence Complex Recommender Systems

Recommender systems recommend movies, restaurants or other items to an active user based on information about users and items in various application domains. Recommendation algorithms are often black box systems in that users can not influence or get explanations on recommended items. However, it is desirable to provide functionalities for users to control the recommendation strategies in more detail, and this is also a requirement of the EU Digital Services Act (DSA). Influencing recommender systems seems more important in complex and high-risk domains such as real estate, travel and tourism, and high-end eCommerce.

The goal of this Master's Thesis in Informatics is to investigate user interaction to influence recommender systems strategies in one or more selected domains. The proposed course of action is as follows:

• Detailed analysis of different recommender systems algorithms and the options to influence their parameters
• Design and prototypical implementation of appropriate user interface elements to control these algorithms
• Setup, conduction, and analysis of a user study to evaluate the effectiveness of the proposed user interface elements

It is not required to implement and study the effect of the control features, thus no backend recommender system. Prerequisites are high motivation, and good enough programming skills and experience in user interface development. Please send your application (brief CV, transcript of records and short motivation statement) to Wolfgang Wörndl (woerndl@cit.tum.de) until March 16th. (Decision and possible start soon after this date.)