For input dataset for DM: We relax the requirement (b) i.e. enumerate join trees (in section 4.1 in ' for DB KWS '). Doing so, the number of intermediate results of the subset of tables shall be much bigger. In other words, the results satisfy the " or " semantics . More result is better for this project because the main focus is to incorporate data mining in KWS. For all the subset of tables, SQL statements are created to retrieve data from the database. The retrieved data is further analyzed using data mining techniques.