International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
Any text mining application may contain side information. This side information may be any links in the document, web logs which contain user access behavior, provenance information, the links for any document or any other non textual attributes which are embedded into the text document. All these attributes may contain a huge amount of information for clustering purposes. But it is difficult to count the concerned importance of this side information especially when some of the data is noisy.