Research on an Interpretable Real-Time Information Recommendation Model based on BAS-ICF algrithm

Most of the existing studies on recommendation system are designed for the accuracy and novelty, but few are designed from the perspective of the user experience, which will cause users to feel confused about the recommendation results, especially in unfamiliar fields. So, this paper will introduce bias into the object-based collaborative filtering (ICF) algorithm. The bias can simultaneously quantify both the quality of the movie and the users' requirements, which not only helps to improve the accuracy of the recommendation, but also enhances the interpretability and real-time performance. It's shown graphically and experimentally. Even though the base-ICF algorithm is not as accurate as the deep-learning algorithm, it performs better in real-time and interpretability, which can achieve a good balance among these three factors.