Research And Implementation Of Mobile Social Networking NG Service Based On Mobile Phone Address Book

SNS is a kind of internet application, which is used to help people establish and maintain social circles in real life. It based on six degree theory and 150 law, provide a platform for users to display, exchange and share information. With the help of SNS, users expand their social circles easily by "friend of a friend" and scientifically manage human network resources. However, after several years of rapid development, almost all SNS web site are faced with the situation that the extent of user participation is low and a large number of users erosion. So how to increase the interaction between users, to ensure a steady relationship between users and website and provide users with more practical social services become the hotspot of the research.This paper presents a mobile SNS system based on mobile phone address book, that build stable and reliable social networks by mobile phone address book. User can access system via PC or mobile phone. In addition to traditional SNS service, system originally provides telecommunication services and personalized value-added services by using user context information.After introducing system structure and each function module, this paper focus on the design and realization of SNS web server. As http protocol cannot adapt to the requirement of system data transmission, we propose an data synchronization mechanism based on flex that can provide two-way interaction between browser and serverExpanding social circles for users is the basic function of SNS. So how to find the potential friends and recommend to the users in extremely large social network is becoming a problem to be solved. To address this issue, we proposed a friend recommendation method based on community division and user similarity in this article. This approach divided the whole social network into several sub-community networks and then computed the similarities between every two users in the sub-community network to find the potential friends that can be recommended to users. Finally, the simulation results show the rationality and effectiveness of this proposed approach.