MA Algorithm to Generate Semantic Web Related Clustered Hierarchy for Keyword Search

Keyword search in XML documents is based on the notion of lowest common ancestors in the labeled trees model of XML documents. It has recently gained a lot of research interest in the database community. In this paper we propose the Modified Active (MA) algorithm which is an improvement over the Active clustering algorithm. The algorithm takes into consideration the entity aspect of the nodes to find the level of the node pertaining to a keyword input by the user. A portion of the Bibliography database is used to experimentally evaluate the Modified Active algorithm. Evaluation results show that MA algorithm generates clusters faster than the Active algorithm and this increases the efficiency of the system