摘要 |
<p><P>PROBLEM TO BE SOLVED: To perform a data mining from a plurality of databases compiled for different purposes by use of data (specific data) largely different in value from other data and having a minor data frequency. <P>SOLUTION: A plurality of data groups are preliminarily associated with each other. For each piece of data included in the data groups, a specific index having a correlation in that they are different in value from other data and have a minor data frequency is obtained, and the specific index is compared with a predetermined reference index. A plurality of data are selected based on this, and the data mining is performed by the selected data. <P>COPYRIGHT: (C)2005,JPO&NCIPI</p> |