Pages 000017
In a huge community site (e.g. "Yahoo!"), it is thought that there is high correlation between the information on the real world and the information acquired in the site. Especially in the community site which has common interest or common concern participants, information in the virtual world and information in the real world are probably closely related. Therefore, it is possible to predict the short-term market expectation and a long-term trend about the related product (or service) by analyzing the logs in the community site. In this research, we analyze the information about a movie. We analyze the transition of the user evaluation on the electronic bulletin board and consider the relation between this information and the information on the real world.