Generating relevant paths using keyword search on compact XML

Abstract

The management of XML data has always been a popular research issue. A simple yet effective way to search in XML database is keyword search. In existing methods, the user has to compose query with which the relevant answers can be retrieved. These methods require the user to have prior knowledge about the data. To overcome the issues arising out of these methods, several approaches have been proposed. In this paper, Two challenges for searching the keyword in XML document has been proposed; 1) how to retrieve high answer semantics matches of the keyword queries (Top-k) 2) how to identify the relevant path for the keyword queries. To identify relevant answers over XML data streams, the Compact Lowest Common Ancestors (CLCAs) are used. We use a compact storage structure (QUICX) system which is efficient both in compression and storage with indexing features for efficient querying. Experiments were carried out using benchmark datasets such as geographical dataset (mondial) and bibliographic dataset (DBLP). In order to prove the effectiveness of the proposed system, it is compared against the existing system with respect to time taken for retrieval and the proposed system achieves about 63.3% of improvement over the keyword search in XML document in terms of time taken for retrieval.