一种融合节点与链接属性的社交网络社区划分算法

针对传统社交网络社区划分算法普遍缺乏对节点属性、链接属性的综合考虑和充分表达利用节点与链接属性信息的模型和机制等问题, 提出了一种融合节点与链接属性的社交网络社区划分算法。该算法融合节点属性的相似度、节点间链接权值等链接属性信息, 定义了相似权值, 并以此为基础, 结合凝聚算法实现了对社交网络的社区划分。实验表明, 该算法对社交网络中属性比较明显的社区划分效果显著。

Combined node and link attributes of social networkcommunity detection algorithm

The traditional social network community detection algorithms generally lack of consideration of node and link attributes, and full expression using node and link attribute information model and mechanism. Aiming at this issue, this paper put forward a kind of combined nodes and links attribute social network community detection algorithm. This algorithm combined with node attributes similarity between nodes, link weights and link information, defined the similar weights, and, on this basis, combining condensation algorithm to realize the social network of community division. Experiments show that effect of this algorithm about social network attribute is remarkable, obviously in attribute-distinct community.