摘要 |
PROBLEM TO BE SOLVED: To provide a method for dividing a hyperlinked database more effectively into hyperilink structure, which are linked by various factors such as the writers, languages and contents of documents. SOLUTION: To separate a desired document from the database including documents having links with other documents, a specific source document from among the documents in the database is determined as a seed document 100 and supplied to a crawler 104. The crawler 104 discriminates, as a cut-set a link such that a document which is linked strongly with the seed document is separated from other documents when removed from the database and removes it to separate a subset 102 that the seed document 104 is strongly connected, from the database.
|