Optimizing XML Data with View Fragments

As web-based applications and data continue to grow, large caches of XML data will result in many application domains. In sensor web applications, there are continuous streams of sensor data being generated, converted to XML and stored for domain queries and data mining purposes. The main problem with these XML caches is that exist-ing XML database queries are very slow, especially for large databases or those with complex structures. In this work we propose a view-based system to XML optimiza-tion where the most popular or well-chosen queries are materialized and fragmented to greatly improve the per-formance of all XML queries