Collaborative filtering recommendation algorithm based on improved user similarity

Abstract

The invention provides a collaborative filtering recommendation algorithm based on improved user similarity, and the algorithm comprises the following steps: S1, obtaining a plurality of scoring items scored by a target user, and selecting a primary user; S2, screening out neighbor users of the target user in the primarily selected users through a basic algorithm to form a neighbor user set; S3, performing score backfilling on the neighbor users of which the scoring items are not scored; S4, obtaining the similarity between the target user and each user in the neighbor user set according to the scoring scores of the target user and each user in the neighbor user set for each scoring item; S5, forming a final neighbor user set by the first k users with the highest similarity with the target user in the neighbor user set; S6, predicting the score of the target user for the new project according to the score of the user in the final neighbor user set for the new project; and S7, performing project recommendation on the target user according to the score of the target user on the new project. And the accuracy of the collaborative filtering recommendation algorithm is improved.

Description

technical field [0001] The invention relates to the technical field of collaborative filtering recommendation algorithms, in particular to a collaborative filtering recommendation algorithm based on improved user similarity. Background technique [0002] As speakers become more and more intelligent, they incorporate recommendation services to provide users with accurate recommendations, so as to improve users' loyalty to products and quality of experience. However, the recommendation algorithms in the existing smart speaker technology ignore the sparsity effect of the user-item rating matrix. The traditional user similarity calculation method is to establish a user-item rating matrix through the user's actual rating records, and then calculate the similarity between users. With the development of the smart speaker industry, the number of users and items has grown exponentially, which has resulted in the user-item scoring matrix becoming extremely sparse. At this time, it i...

AI Technical Summary

Problems solved by technology

With the development of the smart speaker industry, the number of users and items are increasing exponentially, which leads to the extremely sparse user-item scoring matrix

At this time, it is difficult for the traditional user similarity calculation method to calculate the real nearest neighbor set, which leads to a larger prediction error of the score and a decline in the quality of the recommendation.

Images