Music Forum Social Network Mining System Design And Implementation

A social network is a social structure made of nodes (which are generally individuals or organizations) that are tied by one or more specific types of interdependency, such as values, visions, ideas, friendship, trade and so on. A social network always implicates much usable information, so the social network mining technology has been adopted in a variety of fields, including advertising, marketing, computer virus and crime network, terror prediction, technology collaborating etc.. As the times of WEB2.0, the social network mining as a link mining technology has became a new focus of the data mining field. Many academics who are researching social networks have proposed lots of algorithms by means of different ideas to detect community structure of social networks. However those methods are usually applicable to unweighted networks and also cost too much time in the worst case.Nowadays, the network forum is a great channel for netizens to communicate with each other. A network forum based on internet is a virtual space where network users could preconize information, take part in discussion, and exchange opinions. In a word, netizens can participate in the forum freely. The network forum provides a convenient place for people to present their own words. There appears amount of interesting info. As a consequence, a lot of valuable users'information exists in network forums.Among the multitude of forums, music forums have obvious social network features, such as single topics, entire music chain, profession, and steady relations. On this account, this paper will research social networks in music forums. By appropriately defining users'relations as well as topics'relations, firstly we create social network models for a music forum. Secondly we analyze the networks'characters and dig up communities. As regards the network analysis, we make use of SNA (Social Network Analysis) to parse user social network and topic social network, and at the same time a new algorithm for mining forum's key users is presented. In respect of community structure detection, this paper proposes three efficient algorithms aiming at three different weighted networks. These algorithms can effectively dig up user communities with similar preference and hot topic sets. Our research work could support for music forums'all kinds of services, also be of great help to music recommendation. Moreover, it accumulates experience and technology for the social computing based on telecommunications services in the future.