Date Added: Feb 2013
Data mining algorithms search pattern in data. Usually data mining algorithms operate on single table. So pattern is discovered from single table. Mostly data is stored in multiple tables. These tables are related together. By this way the authors can eliminate redundancy and inconsistency. Simple Data mining techniques cannot be applied on this relational data base. For this multi relational data mining is used. Selection graph is used to represent relational learning in pictorial way. In this paper algorithm is presented to translate the SQL query into selection graph.