A novel interest drift sensitivity academic paper recommender based on implicit feedback

Academic recommendation systems have been rapidly developed in recent years, helping researchers to find favorite paper. However, traditional methods applied to paper recommendation face more challenges. First, users can only read a small number of papers, resulting in a very sparse user-paper matrix, but the method based on random sampling of negative samples is ineffective due to the uncertainty of negative samples. And users’ academic interests shift frequently, so the approach that ignores temporal information is not applicable. To overcome the above challenges, this paper proposes an implicit feedback-based interest drift-aware academic paper recommendation algorithm. The algorithm explicitly integrates the user’s interest drift into the model through regularization. The algorithm alleviates sparsity by introducing contextual information through a multiplicative law and significantly reduces the computational complexity by using a caching approach. Experimental results on two real paper recommendation datasets show that the proposed method outperforms current methods in terms of recommendation accuracy and computational efficiency.