Study of Blog Mining for Examination of Tourist Travel Behavior in Japan

The data on how travelers evaluate a specific sightseeing destination are useful for formulating tourism strategies. Making a survey is the most common way to get such data now, but the method has become difficult because travelers have become so diverse that the data are not as meaningful and substantial as they had been. To overcome this problem, Internet blogs are used: comments on sightseeing are so popular that it is easy to collect the needed data. Blog data on tourism were collected in Yatsugatake, Japan, and compared with the data obtained from ordinary questionnaire surveys to examine the features of blog data. Information from 192 blogs was collected. Analyzing the terminology used in the blogs allowed the determination of some attributes of their authors. These attributes were compared with those of questionnaire survey respondents. Although the desired information can be obtained from survey respondents, information from blogs depends on what the authors write about any destination. Nonetheless, because a blogger's destination is clear in his or her blog, the information about destinations can be used by cooccurrence analysis. With categorization of the entry tags of each blog, information about each blogger's interests can be obtained, and their preferences clearly indicate that their destination choices are closely related to their interests. In sum, although the information in blogs is subjective, if used properly the information can be effective in developing strategies for tourism by complementing data from ordinary surveys.