Improvement of pseudo-voting method in recommender systems

Collaborative filtering (CF) is widely used as an engine for recommender systems that identify items preferred by users. The importance of such recommender systems has been recognized in the fields of e-commerce, advertisement systems, and so on. However, the challenge of achieving a trade-off between the "coverage of recommendation" and "accuracy of the predicted score" still exists. Hence, a new recommendation system that utilizes the features of items and the pseudo-voting method is proposed to improve both the coverage and the accuracy. Furthermore, the effect of the proposed system is evaluated in this study.