Academic Journals Database
Disseminating quality controlled scientific knowledge

Clustering Among Multi-Posture Based Correspondence Compute

ADD TO MY LIST
 
Author(s): M. SoumyaHarika, G. Prasadbabu, P. Nirupama

Journal: International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
ISSN 2278-1323

Volume: 1;
Issue: 6;
Start page: 166;
Date: 2012;
VIEW PDF   PDF DOWNLOAD PDF   Download PDF Original page

Keywords: Document clustering | text mining | similarity measure.

ABSTRACT
All clustering models have to presuppose several cluster liaisons among the information substance that they are practical on. Comparison connecting a brace of substance can be distinct moreover unequivocally or unreservedly. In this paper, we bring in a narrative multi-posture based correspondence compute and two allied clustering models. The foremost divergence among a customary divergence/correspondence gauge and ours is that the former uses only a single stance, which is the derivation, while the concluding utilizes many diverse postures, which are substance, implicit to not be in the same cluster with the two substances being deliberate. Using multiple postures, additional revealing judgment of comparison could be achieved. Theoretical investigation and experimental cram are conducted to sustain this claim. Two norm functions for manuscript clustering are projected based on this new gauge. We evaluate them with numerous familiar clustering algorithms that use other admired resemblance procedures on various manuscript collections to validate the compensation of our suggestion.
Why do you need a reservation system?      Affiliate Program