Sorting prediction-oriented social recommendation method

Abstract

The invention discloses a sorting prediction-oriented social recommendation method. The sorting prediction precision is further improved by utilizing concealed features of a user and a project. The method comprises the steps of firstly modeling the potential features of the project and the potential features of the user extracted by using LFM through utilizing a PL model; secondly building a multidimensional trust model by utilizing the user as features of a truster and a trustee; and finally performing joint modeling to build a sorting prediction model, and performing optimization processingon the model to obtain optimal first N recommendation lists. According to the method, while social network structure information is considered, two different roles of the user serve as the truster andthe trustee to perform modeling, so that the social information is combined in recommendation, and the purpose of optimizing a recommendation result can be achieved under the condition of data sparseness.

Description

technical field [0001] The invention relates to the technical field of collaborative filtering recommendation. Specifically, it relates to a ranking prediction-oriented social recommendation method. Background technique [0002] At present, the application of personalized recommendation is divided into two categories: one is rating-oriented recommendation (recommendation is performed using the method of predicting score prediction); the other is ranking-oriented recommendation. The former pays more attention to the problem of rating prediction, assuming that the user's rating data can represent the user's preference, find the highest items in the predicted rating data and recommend them to the user; the latter focuses on the user's choice of each item In terms of possibility, the recommendation list that the user may like is predicted by ranking learning. [0003] Recommendations by learning to rank are an emerging research area. At the 2013 ACM Recommendation Conference,...

AI Technical Summary

Problems solved by technology

However, this method incorporates too little information into the model, and does not consider the impact of contextual information such as item label information, so it affects the accuracy of the recommendation system to a certain extent, and still has certain limitations in practical applications.

Images