Enhancing caching in distributed databases using intelligent polytree representations
In this paper we study the issue of utilizing polytree structures in a real-life application namely that of enhancing caching in distributed databases. Specifically, in this application, the only data or learning cases available is a huge trace of a set of queries of the type of “Select” statements made by different users of a distributed database system. This trace is considered as a sequence containing repeated patterns of queries. The aim is to capture the repeated patterns of queries so as to be able to perform anticipated caching. By introducing the notion of caching, we try to take advantage of performing local accesses rather than remote accesses, because the former significantly reduces the communication time, and thus improves the overall performance of a system. We utilize polytree-based machine learning schemes to detect sequences of repeated queries made to remote databases. Once constructed, such networks can provide insight into probabilistic dependencies that exist among the queries, and thus enhance distributed query optimization.
|Series||Lecture Notes in Computer Science|
Messaouda, O. (Ouerd), Oommen, J, & Matwin, S. (Stan). (2003). Enhancing caching in distributed databases using intelligent polytree representations. In Lecture Notes in Computer Science. doi:10.1007/3-540-44886-1_40