Clustering is one of the data mining techniques which hasimportant utility in real time applications. Cluster is a group of objects with highest similarity. The clustering results can be used further in many applications including query processing. Clustering can also be used in text mining. The existing clustering algorithms in this domain use single viewpoint to find the similarity between object. However, the singe view point similarity measure cannot have highly informative assessment of similarity. In this paper we propose and implement a novel measure known as multiviewpoint based similarity measure. It will consider multipleviewpoints while measuring similarity which facilitates highly informative assessment of similarity. We built a prototype application to demonstrate the proof of concept. The empirical results revealed that the measure is effective.
展开▼
