| 摘要 |
PROBLEM TO BE SOLVED: To recommend suitable contents for current user's taste. SOLUTION: When user's evaluation of K new contents is inputted from a user interface part 11 for M-K old contents, a matrix generation part 18 generates a meta-data matrix a matrix by adding one M-dimensional vector based upon the user evaluation of respective M new and old contents to row components of a matrix consisting of rows corresponding to N meta-data and columns corresponding to M new and old contents. An LSA arithmetic part 20 subjects the meta-data matrix to singular value decomposition to generate its approximate matrix. A content recommendation part 23 uses component values corresponding to user's evaluation of old contents in the approximate matrix to decide whether the old contents can be recommended again. This invention is applicable to an information processor recommending contents. COPYRIGHT: (C)2006,JPO&NCIPI
|
