| 摘要 |
PURPOSE: A method for classifying contents for analyzing data is provided to designate and store contents classification codes capable of analyzing large scaled data such as data warehousing, data mining, user tendency analysis, and a customer relationship management(CRM). CONSTITUTION: Contents are classified horizontally(601) and classified again in vertically according to the horizontally classified contents(602). It is checked whether the vertical classification is completed(603). If the vertical classification is completed, a level graph is prepared(604). It is judged whether the lower vertical node is an AND relation or an OR relation with respect to each horizontal node(605). If the lower vertical node is an AND relation, the number of necessary bit is identified with the number of the vertical classification items(606). The number of necessary bit with each horizontal level of a graph is calculated by adding the number of bit necessary in each horizontal node which constitutes the corresponding level(608), and the number of whole necessary bit is calculated by adding the number of bit desired for each level(609). For storing the contents classification codes, the calculated number of bit is allocated. The bit is allocated from the most significant horizontal level using the level graph per each bit(610).
|
