Towards Efficient Metaquery Generator
Tamar Bash, Rachel Ben‐Eliyahu‐Zohary · Kalpa publications in computing · 2018
Metaquery (MQ) is a datamining tool for inferring relations between data items over a relational database. The concept of MQ leads to autonomous discovery of logical rules involving the database tables. A central module of any MQ system is the MQ generator, which automatically generates all possible MQs to be tested against a database. The MQ generator is supposed to work in an efficient manner, while not missing any meaningful MQ. In this paper we present an algorithm for MQ generation that works as a search algorithm in which the states are all possible MQs, and the search tree is pruned whenever possible. Preliminary experiments prove that, indeed, the approach we take leads to a significant reduction in computation resources.