Optimization of decentralized multi-way join queries over pipelined filtering services
The motivation of this work starts with the need of minimizing query execution time for complex queries performed in a service-oriented infrastructure. Large amounts of data can be found today in every domain such as data resulting as output from various processes like e-commerce transactions, banking transactions or credit card transactions. A part of the process of analyzing very large amounts of data, with emphasis on data mining and machine learning methods, is the combination of different services and data resources. So, finding the order of invocation of the remote services in a query performed over web and grid data infrastructure leads to an efficient execution plan for it. After some motivating examples, authors formulate the theoretical background of the problem and present their \textsc{2 Phase-Service Ordering} algorithm. An enhanced version of this algorithm called \textsc{Advanced 2 Phase-Service Ordering} algorithm is also proposed, which is based on another work of the same authors. Extensive evaluation of the proposed algorithms reveals the situations when each of the algorithms is best suitable.
- Pipelining in multi-query optimization.
- Algorithms for distributional and adversarial pipelined filter ordering problems
- scientific article; zbMATH DE number 970736
- Stochastic optimization for join of three relations in distributed databases. II: Generalization and more applications
- On resource scheduling of multi-join queries in parallel database systems
This page was built for publication: Optimization of decentralized multi-way join queries over pipelined filtering services
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q695027)
