Abstract: The existing social network data partitioning algorithms focus on the social relationship and interaction,without considering location information,which results in the long response time of location-based queries.To solve this problem,we designed a two-layer graph model of mobile social network which takes the location dependency of the user interaction behavior into account.We proposed a mobile SNS data dynamic partitioning and replication algorithm based on location information--MSDPR.MSDPR divides data based on the clustered results generated by an improved K-Means clustering algorithm,and then replicates data by using the social relationships.Experiments reveal that MSDPR can effectively improve the efficiency of the local access and reduce the latency of access in the mobile social network.Moreover,it also has better adaptability when adding data dynamically.

Key words: Mobile social network,Distributed storage,Dynamic partitioning and replication,Location information,User interaction