Research On Point-of-Interest Recommendation Algorithm Based On Multi-Feature Intergration In LBSN

In recent years,with the popularity of mobile intelligent terminals and the rapid development of Internet technology,users usually chose the mobile terminal firstly instead of the traditional PC terminal.The traditional social network is combined with positioning technology to form a Location-Based Social Networks(LBSN).The location-based social networks platform allows users to check-in at their current location anywhere and anytime,and share the checked-in comments with location information to social friends.The platform can use the users' check-in data to find users' hobbies and help users find businesses or places which they are interested in,and at the same time businesses can also use this platform for online promotion.This thesis has conducted related researches on location recommendation in locationbased social networks.The main results are as follows:1.This thesis studies the user's check-in space distribution and proposes a point-ofinterest(POI)recommendation algorithm which is based on user check-in spatial clustering.The algorithm designs a spatial clustering algorithm according to the administrative region information of interest points and the feature of the user's checked-in distribution in the city.Then,according to the distribution of each user's checked-in points of interest in the clustering subsets,the POI recommendation algorithm merged with information such as user preferences,social relationships,and user ratings to recommend points of interest for users.The experimental results show that the improvement strategy proposed in this thesis is effective and improves the accuracy of the recommendation and the scalability of the data.2.In order to solve scalability and the data sparsity in location recommendation,this thesis presents the POI recommendation algorithm that combines the user's interest and geographic factors.Firstly,it combines the adaptive kernel density distribution,naive Bayesian algorithm and the popularity of POIs to explore the user's geographical preferences,and screens out some candidate recommended POIs according to the geographical preference model;then,in order to overcome the data sparsity of collaborative filtering algorithm,the author constructs a user preference model to carry out POI recommendation based on the similarities of user checked-in,category information and user trust;finally,use the Yelp data set to analyze the experiment.The result shows that the POI recommendation model based on user interest and geographical factor obtains a good recommendation effect.