A method for recommending points of interest based on the importance of the points of interest and the authoritativeness of a user

Abstract

A method for recommending points of interest base on the importance of the points of interest and the authoritativeness of a user is disclosed. The method mainly includes the following steps: firstly,according to the check-in behavior of the user and the interest point and the co-occurrence relationship between the interest point and the user, by means of the hybrid model of HITS and PageRank, the user authority and the importance of the interest point are calculated; secondly, the importance of interest points is integrated in the Poisson factor model to model the check-in behavior of users;then, the user authority is integrated into the point of interest recommendation model to measure the confidence of the user's sign-in behavior data; then, the stochastic gradient descent algorithm is used to learn the implicit eigenvectors of users and interest points; finally, according to the implicit eigenvectors of users and interest points, the check-in frequency of unregistered interest points is predicted, and the potential interest points of users are recommended according to the predicted check-in frequency. The method mainly solves the problem of recommending points of interest ina location-based social network.

Description

technical field [0001] The invention relates to the technical field of data mining, in particular to a method for recommending points of interest based on the importance of points of interest and authority of users. Background technique [0002] Recently, location-based social networks (LBSNs) are very popular and have attracted extensive attention from both industry and academia. Typical location-based social networks include Foursquare, Gowalla, Facebook Place, etc. In LBSNs, users can connect with their friends, upload photos, and share their location by checking in points of interest. In addition to providing users with a social platform, LBSNs hope to use users' check-in history and other information to mine user preferences and recommend interesting places that meet user preferences. The task of recommending interesting locations is called POI recommendation. Point-of-interest recommender systems play an important role in LBSNs because they not only meet the individu...

AI Technical Summary

Problems solved by technology

[0003] However, most existing methods simplify user check-in behavior data, that is, use binary values ​​to indicate whether the user has visited a POI no matter how many times the user has checked in to a POI.

In addition, little research work considers the influence of POI importance and user authority on the user's final check-in decision

Images