| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To gather user's preference information by determining the category of a keyword-unspecified HTML document. <P>SOLUTION: A preference information gathering method comprises steps for: generating a keyword database in which category information, a group of nouns belonging to the category, and weights of the nouns in the category are registered; altering registered items of the keyword database by reading an HTML document at each specified period and extracting nouns from the document; extracting nouns from an HTML document that a user has accessed on a Web server; calculating weights of nouns extracted by categories in the categories by reference to the keyword database; deciding the category that the HTML document has been accessed by the user belongs to according to the calculated weights of the nouns extracted by the categories; and storing the decision result in a user preference information database. <P>COPYRIGHT: (C)2005,JPO&NCIPI</p> |
