| 摘要 |
PURPOSE: A product recommendation system is provided to recommend an excellent product to a user by recommending a product in consideration of the product instruction and evaluation information of users who purchase an actual product and the preference of the user who will purchase the product. CONSTITUTION: An evaluation information collector (110) collects the instruction and evaluation information for a specific product. An importance calculator (130) calculates the importance for keywords which are extracted by a keyword extractor (120). An importance personalizing unit (150) calculates the keyword importance by users for keywords which are selected and stored by a keyword selector (140). A group forming unit (160) forms users having a similar importance pattern into a group. A product recommendation unit (170) recommends a product in which a specific user within the group does not purchase among the product in which the users of the formed group purchase to the corresponding specific user. [Reference numerals] (110) Evaluation information collector; (120) Keyword extractor; (130) Importance calculator; (140) Keyword selector; (150) Importance personalizing unit; (160) Group forming unit; (170) Product recommendation unit; (180) User information collecting unit; (AA) Purchase history data per user; (BB) Keyword per product |
