基于PCM聚类算法的Blog社区发现

针对传统的社区发现算法无法发现社区中的核心成员和边界成员的缺点,提出了基于PCM聚类算法的Blog社区发现算法,用来识别Blog社区的核心和边界.首先,使用随机行走的方法计算可以衡量两个Blog亲密度的对称社会距离;然后,在对称社区距离的基础上使用PCM聚类算法对Blog进行聚类,得到每个社区中的成员属于社区的概率表示.最后,通过确定相应的概率阈值,确定社区的核心和边界.实验结果表明:该算法能够获得社区中的成员属于社区的概率,根据这个概率可以确定社区中的核心成员和边界成员.

Blog Community Discovery Based on PCM Clustering Algorithm

Considering that the traditional calculation of community discovery can not find the shortcomings of the core and boundary members of the community,this paper puts forward Blog community discovery algorithm based on soft clustering algorithm PCM to identify the core and boundary of the Blog community. Firstly, the use of calculation with random walk method can measure the symmetrical society distance between two Blogs′ intimacy. Then, on the base of symmetrical society distance, algorithm use PCM to cluster Blog to get the probability of the member in every community group belonging to community group. At last, the core and boundary of the community can be determined through the definition of corresponding probability threshold value. The experiment has shown that the algorithm can obtain the probability of the community member belonging to the community and can find out the core and boundary members of the community according to the probability.