A Context-Aware Collaborative Filtering Algorithm through Identifying Similar Preference Trends in Different Contextual Information

摘要

Three main approaches in Context-Aware Recommender Systems (CARSs) are pre-filtering, post-filtering and contextual modeling. Incorporating contextual information into main process is the different point of contextual modeling from two first approaches. In this paper, we first propose a new context-aware collaborative filtering (CACF) algorithm with contextual modeling approach combined from a clustering technique and matrix factorization method named Similar Trends Identifying (STI). We then compare the proposal with various matrix factorization-based algorithms. Overall, the STI algorithm outperforms some compared algorithms in terms of evaluation metrics and available contextual data sets.