Next-Item Recommendation via Collaborative Filtering with Bidirectional Item Similarity

Exploiting temporal effect has empirically been recognized as a promising way to improve recommendation performance in recent years. In real-world applications, one-class data in the form of (user,...