A Prediction Study on the Localization Policy of Improving the Satisfaction of Chinese Visitors to Japan by Using Matrix Factorization Techniques

Abstract Against the background of the rapid growth of Chinese tourists to Japan, by observing the changes in the patterns of Chinese visitors to Japan and probing into the insufficiency of researches in the past studies, it is believed that online travel reviews can be used to present the current situation of tourism industry and be of value to provide the basic data for long-term sustainable research after analysis, including how to increase Chinese tourists’ satisfaction continuously and how to promote the service level of related industries. In this study, the unknown information caused by the cultural and regional differences in the reviews of Chinese visitors’ to Japan is taken as the prediction object and the local tourists’ travel comments as a reference group. Then, the factor decomposition technology and text mining technology are applied to predict the standards of services and projects that can meet the characteristics and demands of Chinese tourists. The comments of Chinese and Japanese tourists who stayed at hotels in the Hakone scenic area from May 2018 to April 2019 (time span: one year) are taken as the samples in this study. Thus, the disparity in the purpose of travel between Chinese tourists and Japanese local tourists in this scenic spot can be effectively displayed. Besides, it also clearly points out that due to a lack of understanding of Japanese food culture, Chinese visitors were in urgent need of relevant services and assistance. Therefore, this method will provide guidance for the growth of the local tourism industry and also demonstrate the feasibility of research methods. In the end, a complete analysis and calculation method for online comment text mining is established, which also provides guidance on improving the satisfaction of Chinese visitors to Japan. Besides, this method can also be considered to apply in more tourists with the same cultural background.