Personalized event recommendation method and system fusing theme matching and bidirectional preferences

Abstract

The invention discloses a personalized event recommendation method and system fusing theme matching and bidirectional preferences. The method comprises the following steps: firstly, extracting theme information of an event and historical events in which a user participates by utilizing a document theme generation model LDA, and calculating a theme matching degree between the user and the event; secondly, for event-based social network recommendation, considering from the two-way perspective of users and events, constructing preference models of the users and the events, respectively obtaininguser preference scores and event preference scores, and more completely mining preference relationships from the two perspectives of the users and the events; and finally, fusing the user-event pair matching degree with the user-event bidirectional preference linear weighting combination to obtain a final user-event pair comprehensive score, and taking the sorted TOP-K user-event pairs as recommendation results. According to the scheme, the performance of the recommendation algorithm is superior to that of a traditional recommendation scheme, the personalized preference of the user can be wellpredicted, and therefore the purpose of personalized recommendation is achieved.

Description

technical field [0001] The invention relates to the technical field of information recommendation, in particular to a personalized event recommendation method and system that integrates theme matching and bidirectional preferences. Background technique [0002] With the rapid development of the Internet and computer technology, traditional social networks have also developed in different innovation directions in recent years, and some special types of new social networks have been formed, such as location-based social networks (Location-Based Social Network, LBSN), a social network that mainly forms social relationships based on users' geographical check-in information, and another complex heterogeneous social network that combines online and offline - Events-Based Social Network (EBSN), Different from the friend relationship established between acquaintances in traditional social networks, users in event-based social networks establish interpersonal relationships through so...

AI Technical Summary

Problems solved by technology

[0005] Most of the current event-based social network recommendations are mainly based on the user’s one-way point of view to extract feature preferences for recommendation. Although the social influence of the event organizer will be considered, the potential attractiveness of the event is not expressive enough.

On the other hand, regarding the impact of topic factors, only the event topic is considered as one of the recommendation factors, and less consideration is given to the user topic factor and its matching degree with the event topic

Images