Infromation Recommendation
The Information recommendation the technique which presumes the demand of user automatically and recommends the information which fill the demand. The Information Recommend is mainly made by using technology of information filtering, the distinction with an information filtering system. In this text, we call recommendation system which besed on agents, and give a user the filtering effect positively. It is the reason that the word "recommendation" which has active impression and the impression of agents overlaps. To reallize a recommendation system, we need express user's likings on computer.We call this expression user's model(or user's preference), and the process of expressing user modelings. As an available element of structure, content-based filtering which uses keyword vectorsocial filtering which uses user's place or position,we give a few examples. There is "precision" which showed the rate including the information with which a demand of a user is filled while "recall" which showed the rate of the information actually recommended in the evaluation measure of a recommendation system.

Shintani lab. 2003