RECOMMENDATION DEVICE, RECOMMENDATION METHOD, RECOMMENDATION PROGRAM AND RECORDING MEDIUM WITH SAME PROGRAM RECORDED

摘要

PROBLEM TO BE SOLVED: To provide a recommendation device for increasing a customer lifetime value (LTV) regardless of a charging model.

SOLUTION: The recommendation device 1 is provided with a pre-processing part 21 for creating input data 47 by extracting the purchase history of merchandise for every user by using a purchase log 46; a post-charge rise probability estimating part 22 for estimating post-purchase rise probability 48 showing probability that LTV rises according to purchase with a purchase frequency model by using the input data 47; a post-recommend purchase probability estimation part 23 for estimating post-recommend purchase probability 49 under the consideration of the taste of a user by using the input data 48 and the degree of influence of recommend; a post-recommend rise probability estimation part 24 for estimating the post-recommend rise probability 410 showing probability that the LTV rises according to recommend by using the post-purchase rise probability 48 and the post-recommend purchase probability 49; and a recommend part 25 for selecting merchandise whose post-recommend rise probability 410 becomes the maximum, and for presenting it as the object of recommend.

COPYRIGHT: (C)2009,JPO&INPIT