Comprehensive similarity migration-based collaborative filtering algorithm

Abstract

The invention discloses a comprehensive similarity migration-based collaborative filtering algorithm. According to the algorithm of the invention, in the aspect of similarity calculation, user ratinginformation and user attribute information are both utilized, satisfaction degree scoring criterion differences between users are considered, a method which adopts the distribution consistency of userratings to measure the similarity of the user ratings is adopted, and therefore the comprehensive similarity migration-based collaborative filtering algorithm can improve the accuracy of similarity calculation and improve the quality of data migration compared with the prior art. The experimental results show that the model can effectively alleviate a data sparsity problem compared with other algorithms. In the future, the data of auxiliary fields can be migrated through combining the similarity of projects or other knowledge such as text information, and therefore, recommendation accuracy can be improved.

Description

technical field [0001] The invention relates to the field of network information calculation, in particular to a collaborative filtering algorithm based on comprehensive similarity migration. Background technique [0002] At present, the amount of network information is increasing exponentially. On the one hand, network users can obtain a wealth of information, but on the other hand, they are faced with the problem of information overload, and it is difficult to mine useful information from the massive information. The recommendation system can filter out the parts that the user is interested in from massive data according to the user's interests. At present, recommender systems have been widely used in e-commerce platforms such as Amazon, eBay, MovieLens, and GroupLens. [0003] Collaborative filtering technology is one of the most widely used technologies in the recommendation system. Its basic idea is to use the user's historical rating data to predict the user's interes...

AI Technical Summary

Problems solved by technology

[0007] Although the above algorithms all use auxiliary domain knowledge to improve the recommendation accuracy, there are still some shortcomings: first, the model based on matrix transformation has more model training parameters; The third is that when calculating the similarity of user ratings, the differences in scoring standards for user satisfaction are ignored

The existing similarity-based migration models generally only use user rating information, and ignore the difference in user rating standards in the calculation of rating similarity

Images