Social network friend recommendation system and method based on activity similarity and social trust

Abstract

The present invention is a social network friend recommendation system and method based on activity similarity and social trust, which belongs to the field of information recommendation and data mining. The method mainly uses user social trust value and activity preference similarity to implement friend recommendation in location-based social networks. , since activities can reflect user interests and preferences, friends with similar preferences can be found through the similarity of activities between users; since social trust can reflect the closeness of interactions between users, friend recommendations based on different levels of trust relationships are more reasonable and interpretable. ; Experiments have proved that the recommendation effect of the present invention is better than the existing friend recommendation method in terms of accuracy and reasonable interpretability, and the practical application value is very high. If it can be promoted, it can clarify the scope of target customers for enterprises and institutions and improve the relevance of advertising services. Degree and accuracy, improving the value of advertising and marketing have important guidance and decision-making significance.

Description

technical field [0001] The invention belongs to the field of information recommendation and data mining, and in particular relates to a social network friend recommendation system and method based on activity similarity and social trust. Background technique [0002] With the increasing prosperity and development of social networks, almost the whole world has been covered by social networks, among which location-based social network services LBSNs (Location-based Social Networks) are the most attractive. LBSNs can strengthen the correlation between social network and geographical location by recording when and where users have taken place, so as to provide various marketing decision-making services for enterprises. [0003] Friend recommendation is one of the many services of LBSNs, and it is also one of the current research hotspots in academia and industry. However, according to relevant information, so far, most of the friend recommendation methods in LBSNs are based on l...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problems of false recommendations and unconsidered user behavior preferences in current friend recommendation methods, the present invention proposes a social network friend recommendation system and method based on activity similarity and social trust, referred to as FRBTA (Friend Recommendation Based on Trust and Activity) method, to achieve The purpose of improving practical value, clarifying the scope of target customers, improving the relevance and accuracy of advertising services, and increasing the value of advertising and marketing

Images