抄録
In these days, people expect functions to comprehend characteristics of the user, for example interest and liking, to communicate with computers smoothly. In this study, we constructed a system to visualize an interest model of the user based on information in browsed Web pages by the user by extracting words and interword relationships. In this model, metadata is appended to words and interword relationships. Kinds of metadata of words are personal name, corporate name, site name, name of commodity, product name and location name. Metadata of interword relationships is prepared to these words. This system makes a map by visualizing this model. This system has functions to modify this map. We showed efficacy of this system by using the evaluation experiment.
