| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To detect a user's potential concern through the user's daily unintentional information processing behavior to present a concrete stock matching the user's potential concern. <P>SOLUTION: A stock recommendation system 10 includes: a user data collection part 12 which extracts a keyword from data inputted by a user or data browsed by the user and stores the keyword in a user keyword storage part 16 in association with a user ID and a date and hour; a company attribute storage part 28 wherein attribute information such as locations, types of business, and products of publicly-traded companies is stored; a keyword selection part 18 which refers to the user keyword storage part 16 to totalize frequencies of respective keywords within a set period for every user, and selects the prescribed number of keywords on the basis of the frequencies; and a recommendation stock part 26 which refers to the company attribute storage part 28 to specify companies having the respective selected keywords as an attribute. <P>COPYRIGHT: (C)2011,JPO&INPIT</p> |
