Georgios Klioumis, "A Novel social choice mechanisms-based recommender system for the movies domain", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2024
https://doi.org/10.26233/heallink.tuc.98707
This is a recommender system for the movies domain. Employing novel item and user probabilistic modelling techniques. Movies are modeled based on their summaries with various Natural Language Processing techniques. Our recommendation process uses a social choice theory mechanism, based onmulti-winner elections. Our results indicate that (i)movie summaries are indeed a valuable tool for movie classification, (ii)recommender systems can benefit from the use of probabilistic modelling, (iii)different user types can benefit from different recommendation approaches, (iv)our movie recommendation approach using social choice theory mechanisms is effective on real-world data.