Discovery from Queries

Stronger integration of database technology and data mining has been identified as one of the most important directions in knowledge discovery from databases (KDD) research. Database technology can certainly contribute more than just data preparation for KDD, and there should be ways to further explore new information from data processing. This paper discusses the possibility of using query results from databases with existing DBMSs as a good source of information for forming hypotheses for KDD, and designs a prototype system, called DQ (Discovery from Queries), to demonstrate its feasibility.