Research And Implementation On POI Recommendation System Based On Embedding Method

As the increasing popularity of the applications of location-based service,locationbased social networks(LBSN)have attracted a large number of users to Check-in at their preferred points-of-interest(POI)and share their experience and the fun of visiting these POI with their friends.As a result,a large number of POI check-in data which have location information and social network information was generated.In order to extract valuable information from these check-in data and provide more user-friendly services for users,the technology of the POI recommendation has emerged.The POI recommendation has a great value to help users explore the surrounding living environment and improve the quality of life.Therefore,POI recommendation has received extensive attention from the academic community and industry,currently,and has become a hot-spot in the recommended field.This paper mainly focuses on the POI recommendation.Our paper uses the based on embedding method to describe the relationship between POI and users,POI and POI in a more fine-grained describe to provide higher-quality personalized recommendation services.The main research works in this paper are as follows:(1)This paper focuses on the POI recommendation problem.We introduce the existing POI recommendation technologies,and describe the representative POI recommendation algorithms,analyze existing problems in those algorithms.This provide a strong theoretical support for the follow-up research work.(2)Against the existing problems in the social network of users,we propose a social personalized embedding model which is based on embedding method.And we used more accurate user social network relationship to improve the POI recommendation accuracy.(3)To solve the problem of different attributes of POI,we propose a hierarchical embedding model of POI attributes based on embedding method.This model uses the embedding method to accurately describe the attribute relationships between POI and POI,which are used for improve the performance of POI recommendation.(4)This paper illustrates the effectiveness of our proposed methods in the real-world dataset.The experimental results show that our proposed the POI recommendation algorithms based on embedding method in this paper is outperforms the state-of-the-art POI recommendation algorithms in terms of several evaluation metrics.In addition,this paper also implements a POI recommendation system based on embedding method.