Approaches to Ontology Based Algorithms for Clustering Text Documents V.Sureka S.C.Punitha

The advancement in digital technology and World Wide Web has increased the usage of digital documents being used for various purposes like e-publishing, digital library. Increase in number of text documents requires efficient techniques that can help during searching and retrieval. Document clustering is one such technique which automatically organizes text documents into meaningful groups. This paper compares the performance of enhanced ontological algorithms based on K-Means and DBScan clustering. Ontology is introduced by using a concept weight which is calculated by considering the correlation coefficient of the word and probability of concept. Various experiments were conducted during performance evaluation and the results showed that the inclusion of ontology increased the efficiency of clustering and the performance of ontology-based DBScan algorithm is better than the ontology-based K-Means algorithm