期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2013
卷号：4
期号：3
页码：480-482
语种：English
出版社：Ayushmaan Technologies
摘要：Clustering method in similarity measure space is a topic for new research graduates. In this paper a new spectral clustering method (CPI- Spectral Clustering Method has been introduced. Here the documents are expected into a low-dimensional semantic space in which the correlations between the documents in the local patches are maximized and minimized simultaneously. Correlation as a similarity measure is more suitable for detecting the intrinsic geometrical structure of the document as it embedded in the document. CPI can discover easily the intrinsic structure as compare to the Euclidean distance. We have demonstrated the result in rigorous experiment and shown that CPI method is better than the existing one.
