Correlation Preserving Indexing Based Text Clustering
Venkata Ramana S · IOSR Journal of Computer Engineering · 2013
In Document clustering previously they presented new document clustering method based on correlation preserving indexing.It simultaneously maximizes the correlation between the documents in the local patches and minimizes the correlation between the documents outside these patches.Consequently, a low dimensional semantic subspace is derived where the documents corresponding to the same semantics are close to each other with learning level parsing procedure based CPI method.The proposed CPI method with learning level parsing procedure is to find correlation between relational documents to avoid maximum unknown clusters those are not effectual to find exact correlation between documents depend on accuracy of sentences.The proposed CPI method with learning level parsing procedure in document clustering doubles the accuracy of previous correlation coefficient.The proposed hierarchical clustering algorithm behavior is different with CPI in terms of NMI, Accuracy.