O- Recommend: An Optimized User-Based Collaborative Filtering Recommendation System
Lei Zhang, Xuan Liu, Yidi Cao, Bin Wu · 2018
When people purchase products on the Internet, the overwhelming information makes it difficult to choose a satisfactory merchandise. Hence, an effective recommendation system seems to be very necessary. The user-based collaborative filtering recommendation is the earliest and most popular recommendation system. The most significant step of user-based collaborative filtering recommendation is comprehensive user similarity calculation. However, most recommendation systems ignore the indispensability of user evaluation normalization and the weighted user attributes in comprehensive user similarity calculation, which leads to the inaccurate recommendation. Based on these issues, this paper proposes an optimized user-based collaborative filtering recommendation system, called O-Recommend. O-Recommend not only validates the necessity of the user evaluation normalization and the weighted user attributes in the comprehensive user similarity calculation, but also improves the recommendation accuracy.