Research On Recommendation Of Friends And Venues In The LBSN

With the popularization of location services, location based social networks (LBSN) are becoming increasingly popular and affecting people’s daily life. Friends and venues are the two main themes in the LBSN. The recommendation of friends and venues has become an important issue which attracts attention by both academic and LBSN platforms. So the algorithm of friends and venues recommendation in the LBSN is the core of this paper.To improve the friend recommendation performance in the LBSN, a kind of friend recommendation algorithm considering users’preference of making friends is proposed in this paper. Users in LBSN ware described by interest similarity attribute, distance attribute and familiarity attribute. The interest similarity attribute was computed based on information entropy theory. The distance attribute was computed based on the Naive Bayesian. The familiarity attribute was computed based on the quantity of same friends. Users’ preference was reflected by weighting coefficient of the three attributes which comes from the friend-list of the user. To improve the venue recommendation performance in the LBSN, a kind of venue recommendation algorithm combined with similarity and friend-trust is proposed. Users were divided into non-friends set and friends set for target the user. The algorithm based on similarity was designed for non-friends set, and the algorithm based on friends’trust was designed for friends-set. The similarity based algorithm was built on material diffusion theory, and optimized by friendship. The friends’trust based algorithm was built on social influence factor.The contrast experiment was designed and conducted on the real dataset of Foursquare. Three indicators including precision, recall and Fl-measure ware selected to evaluate the algorithm. The results show that the two algorithms proposed in this paper achieves good performance in different evaluating indicators.