| 摘要 |
PROBLEM TO BE SOLVED: To provide an item recommendation device capable of recommending an item related to the topic of a subject of conversation based on a trend of the world, and related to a program.SOLUTION: An item recommendation device includes: a topic extraction part 20 for extracting a topic including a feature word from the group of texts; trend store calculation means (30, 40) for calculating and storing the trend score of the topic; program-related Web information extraction means (60, 70) for extracting and storing Web information related to program information; a program/topic relevancy degree calculation part 80 for calculating a relevancy degree between a program and the topic on the basis of program information, program-related Web information and the topic; a program/item relevancy degree calculation part 100 for calculating a relevancy degree between the program and the item on the basis of the item of an item information DB 90, the relevancy degree between the program and the topic and the trend score; and an output part 110 for outputting an item to be recommended in the program on the basis of the relevancy degree between the program and the item. |
