XML pub/sub systems have recently emerged as application-layer XML routers. XML-based network traffic is expected to have significant growth. The main problem of the state-of-the-art XML routing schemes is that XML filtering speed often cannot match the XML document arrival speed. The XML filtering speed is propotional to the number of quries which are typically represented as XPath quries. Aggregation is a useful technique for reducing the number of XPath queries. This paper will present a new XPath query aggregation algorithm based on a node region encoding scheme which provides positional information. Compared with the existing aggregation algorithms in the literature, our proposed algorithm can efficiently evaluate the ancestor-descendant and parent-child relationships between any pair of nodes in XPath queries and process a tree-structured query as a unit. Experimental results demonstrate the effectiveness of the proposed algorithm. The performance improvement for the proposed algorithm could be up to 61% compared to the existing XSearch algorithm which is the most efficient algorithm for XPath query aggregation so far.

Additional Metadata
Keywords content routing, homomorphism, publish/subscribe, query aggregation, region encoding
Persistent URL dx.doi.org/10.1109/SAINT.2011.14
Conference 11th IEEE/IPSJ International Symposium on Applications and the Internet, SAINT 2011
Cao, Y. (Yang), Lung, C.H, & Majumdar, S. (2011). An XPath query aggregation algorithm using a region encoding. Presented at the 11th IEEE/IPSJ International Symposium on Applications and the Internet, SAINT 2011. doi:10.1109/SAINT.2011.14