LBSN Application Related Service Recommendation And Incentive Mechanism

In recent years,with the rapid development of computer network,people's social way gradually from face-to-face communication to social network.At the same time with the positioning technology continues to progress,the traditional social network is the location-based social network(Location Based Social Network,LBSN)direction.The user position information in LBSN has become one of the social attribute data,researchers with the help of the user's social attributes,understand the user's personal preferences and behavior,to provide a series of personalized recommendation for users,which makes LBSN by business and academic wide Related services.LBSN recommend more and more attention,and points of interest(Point Of,Interest,POI)is recommended as a personalized recommendation in LBSN related services is also of concern.However,due to the LBSN data is extremely sparse,resulting in many algorithms recommendation accuracy is not high.Therefore,the analysis of the research status and deficiency the associated LBSN service recommendation in a point of interest and a recommendation method recommended by friends online users based on the incentive method,the main results include:Presents a recommendation algorithm based on user activity zoning meta path based on the interest points.Firstly,analysis of user sign points of interest,showing the regional characteristics according to the users and comments of the place,the user activity area is divided into active regions and non frequent activities in the region;and then according to the structure characteristics of LBSN construction of the two quadrant model between the user activity area and activity area-points of interest;secondly introduce meta path,correlation calculation path to interest from the user point;finally,according to the correlation degree to generate recommendation list.The results show that the algorithm is better than the traditional LBSN algorithm has better recommendation accuracy and recall The recommended rate,the effect is better.Put forward a kind of incentive method recommended by friends online based on users.This method through the analysis of the same area users have more similar sign points of interest,at the same time,there are more tastes easier for users to become friends.This method first analyzes the history of user sign points of interest,the user activity area is divided into frequent activities in the region and the non frequent activities in the region;and then introduce the meta path,according to the users interest correlation,recommended to the user.At the same time,this paper will be the friends of friends recommended mechanism and virtual points based incentive mechanism,in order to better achieve the user incentive.