Prediction of Eudaimonic and Hedonic Movie Characteristics From Subtitles Elham Motamedi 1 , Marko Tkalcic 1 1 University of Primorska, Koper 6000, Slovenia Abstract Personalization and explaining the recommendations of recommender systems (RSs) have recently gotten more attention from researchers. We would like to use eudaimonic and hedonic perceptions of users from movies to personalize the movie recommendations. To achieve this goal, frst, we need to predict the movies’ eudaimonic/hedonic quality features. This research study presents the pipeline and preliminary results for predicting movies’ eudaimonic and hedonic characteristics. Keywords personalization, recommender system, eudaimonic perception, hedonic perception 1. Introduction Movie recommender systems have been studied extensively in the last two decades [1]. Roughly, we distinguish three groups of algorithmic approaches: (i) content-based (CBR), where the recommendations are based on item characteristics, (ii) collaborative, where the predictions are based on similarities between users and items [2] and (iii) hybrid, where the recommendations are based on item characteristic and similarities between users and items. This work addresses CBR. The proposed idea can also be exploited in hybrid approaches. Related work has exploited a wide variety of item characteristics for movie recommendations, such as genre, year, actors, directors, etc. While these characteristics have been proven to contain information about user preferences, we conjecture that item characteristics that have a more intimate refection of the user perception of the item might carry more information about user preferences, hence leading to better recommendations. We aim at investigating a novel set of item characteristics, namely the hedonic and eudaimonic qualities of a movie. The eudaimonic quality is related to the meaning or the goal of the life that one pursues, while the hedonic quality is more associated with the plain pleasure that one experiences [3, 4, 5]. The more meaningful a multimedia item is for a user, the higher the value of eudaimonic perception is for that person. Similarly, the hedonic perception score is the degree of plain pleasure that the user experience while consuming the multimedia item. Therefore, each multimedia item and in particular each movie, can get a score of eudaimonic/hedonic perceptions for each user. Each item has a distribution of eudaimonic/ hedonic perceptions of users. We have frst tried to predict the average eudaimonic/hedonic qualities for movies. However, for future work, we can try to model users and predict for each Human-Computer Interaction Slovenia 2021, November 11, 2021, Koper, Slovenia E elham.motamedi@famnit.upr.si (E. Motamedi); marko.tkalcic@famnit.upr.si (M. Tkalcic) O 0000-0003-0127-4997 (E. Motamedi); 0000-0002-0831-5512 (M. Tkalcic) © 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Workshop Proceedings (CEUR-WS.org)