社会网络中基于随机游走的名称消歧算法

Random walk based name disambiguation algorithm in social networks

在社会网络中,名称对实体的正确引用是对网络进行推导和分析而得出正确结论的基础。为了消除社会网络中名称对实体引用的歧义,基于网络中的随机游走思想提出一种名称消岐算法。首先,根据各个名称在源中的同现情况构建社会歧义网络。然后,在社会歧义网络中,基于随机游走模型计算出歧义节点之间的相似度。最后,采用层次聚类方法对歧义节点进行分组从而实现名称的消岐。实验表明,提出的算法与相关的名称消岐算法相比不仅准确性高,而且执行效率快。

In social networks, reference correctness of name to entities is the basis for correct inference and analysis in the network. In order to disambiguate reference of name to entities, this paper proposed a name disambiguation algorithm based on random walk in social networks. Firstly, it constructed social ambiguous networks according to the co-occurrence of names in a source. Secondly, it computed the similarity between ambiguous nodes based on random walk in social ambiguous network. Finally, it implemented the name disambiguation algorithm by hierarchical clustering. The experiments show that, the proposed algorithm has better accuracy and higher execution efficiency than related works.