摘要 |
PROBLEM TO BE SOLVED: To solve the problem that data representing each cluster can not properly be determined with respect to a data set that is hierarchically clustered.SOLUTION: This data classification device includes: a lowermost layer representative data determining means 3 for respectively determining data representing each cluster about each cluster belonging to the lowermost layer in a hierarchical cluster structure that hierarchically clusters each data included in a data set; and a hierarchical representative data determining means 4 for determining data representing a cluster having high centricity among respective clusters among the respective data determined by the lowermost layer representative data determining means 3 as data representing a cluster belonging to a layer higher than the lowermost layer by one, and further repeatingly performing processing for determining that data representing a cluster having high centricity belonging to a layer to be processed as data representing a cluster belonging to a layer higher than the layer to be processed by one until reaching the uppermost layer. |