Multi View Point Similarity Measure With Incremental Clustering
Shalini Chowdary, S. Priyanka Chadalavada, Sir C. R. Reddy · 2013
The main task of Clustering technique in data mining and text mining is to analyze datasets by dividing it into meaningful groups. certain relationship exists between the objects in the dataset. The similarity between the data objects can be described either explicitily or implicitily. In the existing algorithm the similarity and dissimilarity between the objects is measured using single view point, which is the origin. The drawback is that the clusters can’t exhibit the entire set of relationships among objects. To overcome the conflict, in this paper a new measure of similarity called multiview point based similarity measure is proposed.This approach makes use of different viewpoints,in clustering the web documents which show all relationships among objects where more informative assessment of similarity could be achieved. Analysis and experimental study are conducted in support of this approach. several well-known clustering algorithms that use other popular similarity measures on various document collections are compared to verify the advantages of our proposal.