Collaborative Filtering Recommendation Algorithm Based on Social Relation and Geographic Information
Dongni Ma, Liyan Dong, Kelu Li · Proceedings of the 2nd International Conference on Computer Science and Application Engineering · 2018
Collaborative1 filtering algorithm has such problems as sparsity of raw data, low efficiency and accuracy of recommendation. To solve the problems, this paper proposed the collaborative filtering algorithm based on complementary conditions of social relationship and geographic information. The algorithm firstly introduced the social relation data into the matrix complementation process, which reduces the sparseness of the original user item scoring matrix and enhances the authenticity of the complement data; then used the user's geographic information to filter the information that is used to complement matrix, which drops the error of complementing data and improves the accuracy of complementing data; finally, selected complement items conditionally, which increases the recommendation efficiency and recommendation accuracy of the algorithm remarkably. The algorithm is verified through experiments, and the experimental results prove that the improved algorithm is correct and effective in solving the problem of low original data sparseness and low recommendation accuracy.