Tezt Clustering Algorithm Based on Spectral Graph Seriation
In the field of information processing, most of the existing text clustering algorithm is based on Vector Space Model(VSM). However, VSM can not effectively express the structure of the text so that it can not fully express the semantic information of the text. In order to improve the ability of expression in the semantic information, this paper presents a new text structure graph model. With the weighted graph, this model expresses the characteristics term of the text and its associated location information. On this basis of spectral graph seriation, a spectral clustering algorithm is put forward. This algorithm replace solving common subgraph with matrix computation, then reduce the computational complexity of graph clustering. There are also algorithm analysis and experiment in the paper. The results of the study show that the text clustering algorithm based on spectral graph seriation is effective and feasible.
Graph Model Spectral Graph Theory Tezt Clustering
Guo Wensheng Li Guohe
Department of Computer Science and Technology, China University of Petroleum-Beijing, Beijing Changp Department of Computer Science and Technology, China University of Petroleum-Beijing, Beijing Changp
国际会议
2009年中国控制与决策会议(2009 Chinese Control and Decision Conference)
广西桂林
英文
4255-4259
2009-06-17（万方平台首次上网日期，不代表论文的发表时间）