Similarity Based Clustering with Indexing for Semi-Structured Document

Abstract

Problem statement: To improve the performance of data retrieval in a homogeneous large XML document. Approach: Clustering of XML elements based on the content with indexing. The element which is used for clustering has been identified from the document and/or XML schema. This element is used as a parameter for clustering. The suitable index is created after clustering. Results: The clustering combined with indexing strategy support the efficient retrieval of XML element from the document. Conclusion: The proposed method is used to improve the efficiency of XML data manipulation and comparatively give the better performance rather than clustering or indexing alone.