| 摘要 |
The present invention relates to a purchase recommendation service system and method. The purchase recommendation system comprises: a purchase history database (DB) storing two or more users and purchase history including purchase item and a purchase date of each user; a periodical purchase history extracting device dividing an arbitrarily set entire purchase period into N unit purchase periods and extracting purchase history of each unit purchase period; and a collaborative filtering processing device calculating similarity of users or item preference of users during the N unit purchase periods to obtain an average value during the entire purchase period, and calculating item preference of each user according to the obtained average value to create recommendation information in order of preference. Thus, since purchase history of each user is divided into detailed period units and recommendation information is created to reflect the latest purchase history and a relatively old purchase history with different levels of importance, reliable recommendation information that is trendy, while reflecting past purchase history, can be provided. |
