|International Journal of Applied Information Systems|
|Foundation of Computer Science (FCS), NY, USA|
|Volume 5 - Number 6|
|Year of Publication: 2013|
|Authors: Anju Vijayan, V. Uma, G. Aghila|
Anju Vijayan, V. Uma, G. Aghila . Frequent XML Query Caching in ebXML based E-commerce Applications using Association Rule Mining. International Journal of Applied Information Systems. 5, 6 ( April 2013), 8-11. DOI=10.5120/ijais13-450924
Mining frequent XML query patterns and caching the results can improve the performance of XML based systems. Temporal features of queries can be used to guide cache replacement. Many approaches have been proposed to mine frequent XML query patterns and caching the results. Those approaches mine frequent subquery patterns from historical queries. But for ebXML (electronic business using XML) based applications in e-commerce, most of the queries will be of same structure. In such cases instead of mining subqueries, frequent queries can be mined directly. In this work an efficient system for frequent XML query caching in ebXML based e-commerce applications is proposed which incorporates temporal features of frequent queries for cache replacement. Association rules are mined between frequent queries to discover temporal patterns among them. Semantically closer infrequent queries are clustered and their associations with frequent queries are also mined. These rules are used to guide cache replacement so that subsequent query results will not get replaced from cache.