Personalized Recommendation Research For Location-based Social Networks

In recent years,with the rapid development of Internet technology,the network positioning technology has been greatly improved.The location-based social media has appeared in the Internet social network.Location-based social networking can obtain the user's location information and friends' location information,etc.Promote the development of social networks.In the life,more and more users use the social platform to share and obtain location information.As the number of users' share increases,social media has accumulated a lot of user location information.How to effectively use social positioning information becomes a focus of researchers.Many scholars have carried out research on personalized recommendations based on geographic locations.In the course of using social networks,people generate a large amount of data,which contains a large amount of user information,including user behavior information and preference information,which can be used to provide users with personalized location-based services,such as interest.Point(POI)recommendation.Effectively using the user's location information to specifically and accurately make real-time user recommendations can meet the needs of users,but also greatly increase the user's activity in social networks,so that users more involved in social network.Compared with the traditional online social network,the location-based social network can more accurately help the recommendation system to discover the user's preferences and interests after adding the geographical location information.In many current social location recommendation systems,there are mainly friend recommendation,activity recommendation,and location recommendation.Among these,location recommendation has become a focus of research.Currently,the cold start problem has always been an unavoidable issue in the study of many location-based recommendation methods.This paper proposes a new recommendation mechanism,based on regional active users to recommend,calculate the active user in the user's area through the user's location information,and then recommend the active user's check-in location in the area to the user.This recommendation mechanism solves the cold start problem to some extent.In the research of location recommendation algorithm,many recommendation algorithms do not consider the time factor,and time factor is an important consideration in the recommendation process.The recommendation method proposed in this paper not only considers the location information of the user at that time,but also considers the time information recommended for the user at that time,and the recommendation is more timely and accurate.At the same time,this article takes into account two important factors in the location recommendation,one is the popularity of the location and the other is the influence of the friendship.The position popularity based location recommendation and the friend relationship based location recommendation were studied separately.Finally,a mixed model based on the recommendation of the regional active user,the recommendation based on the popularity of the location,and the location recommendation based on the friend relationship is combined to make the recommendation effect better.In this paper,the algorithm proposed by the paper is verified through specific experiments.The effects of the algorithm are evaluated by Precision,Recall and FMeasure.The experimental results show that the proposed algorithm has a good recommendation effect.