A Hybrid Video Recommendation System: Prioritizing Features Using AHP

In the era of accelerating technological development, the importance of mental health is declining significantly. The impact of our daily content intake on emotional well-being is clearly visible. However, one needs to watch videos carefully because their consumption may lead to more depression. It is necessary to filter out the contents that can help to lighten up our mood. Therefore, in this work, we present a novel recommendation system that combines collaborative filtering with emotional analysis of videos to enhance the accuracy and relevance of content suggestions. The system combines two approaches: collaborative filtering to assess similarities between users and their preferences and extracting emotions from video text with the help of Natural Language Processing (NLP). By fusing these methods, we are able to retrieve both contextual relevance and emotional resonance of content. As a way to further rectify our recommendations, we apply the Analytic Hierarchy Process (AHP), which combines the results and ranks them to ensure balanced and effective suggestions that can gradually improve someone’s mood.