| 摘要 |
PROBLEM TO BE SOLVED: To attain high-speed and high-accuracy Know-Who retrieval capable of calculating individual's knowledge information for one-mail or Web content from the use tendency thereof to be used for retrieval. SOLUTION: The Know-Who retrieval is attained by including steps of extracting a characteristic character string contained in the mail or Web content; extracting, based on the extracted characteristic character string, an individual knowledge attribute formed by comparing appearance frequency information of the characteristic character string in a group of members containing the former of the mail or Web content; forming profile data of the preparer based on a group of extracted knowledge attributes to thereby calculate and accumulate individual knowledge information; calculating a knowledgeable degree of a person having knowledge designated by a Know-Who retriever using the accumulated profile data; and displaying a retrieval result using the calculated knowledgeable degree. COPYRIGHT: (C)2009,JPO&INPIT
|
