Collaborative Filtering Recommendation Algorithm Based on User Explicit Preference
Bo Hu, Zhaohua Long · 2021
Collaborative filtering algorithm is one of the widely used methods in recommender system, However, collaborative filtering has the problem of data sparsity and the calculation of similarity is simple relatively, which can't describe the similarity between users well. Considering that user's preferences to projects may be affected by project attributes, this paper will introduce project attribute information to mine the user's attribute preference and makes up for the problem in similarity calculation by the user's ratings. By describing the user's attribute preference, we propose a collaborative filtering algorithm CF-UEP based on the user's explicit attribute preference. Through experimental comparison, we prove that our method is more advanced. The algorithm solves the data sparse problem effectively.