A Book Recommendation System Considering Contents and Emotions of User Interests

Although the benefits of reading are widely recognized, many people seldom read even though they often claim to have interest in reading. Since conventional book recommendation systems require keywords or a browsing history related to books that reflect user interests, users who rarely read struggle to obtain satisfactory results. In this study, we propose a book recommendation system that enables both users who read habitually and those who rarely read to easily get results that reflect their interests with their own content of interest as queries. Our proposed method identifies recommended books based on the similarity of the vectors of contents and emotions, contained in tweets about the content of user interests and book reviews. In this study’s experiments, we confirmed the effectiveness of our proposed method.