Learning Query-transformation rules are vital for the success of semantic query optimization in domains where the user cannot provide a comprehensive set of integrity constraints. Finding these rules is a discovery task because of the lack of targets. Previous approaches to learning query-transformation rules have been based on analyzing past queries. We propose a new approach to learning query-transformation rules based on analyzing the existing data in the database. This paper describes a framework and a closure algorithm for learning rules from a given data distribution. We characterize the correctness, completeness, and complexity of the proposed algorithm and provide a detailed example to illustrate the framework.
|Original language||English (US)|
|Number of pages||15|
|Journal||IEEE Transactions on Knowledge and Data Engineering|
|State||Published - Dec 1993|
Bibliographical noteFunding Information:
Manuscript received October 1, 1992; revised June 15, 1993. This work was supported by the Graduate School of the University of Minnesota and the Minnesota Department of Transportation. S. Shekhar, A. Kohli, and M. Coyle are with the Department of Computer Science, University of Minnesota, Minneapolis, MN 55414. B. Hamidzadeh is with the Department of Computer Science, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong. IEEE Log Number 9212795.
- discovery in databases
- rule discovery
- semantic query optimization