Comparative Topic Analysis of Japanese and Chinese Bloggers

This paper first studies how to apply a topic model to Chinese and Japanese blog posts collected from a few hundred Chinese and Japanese bloggers and then to classify bloggers into topics. The estimated topics are exploited in the task of over viewing the Chinese and Japanese bloggers' concerns, opinions, and cultures. Those topics are also quite helpful when comparing them between Chinese and Japanese in order to discover differences in the concerns, opinions, and cultures of the two languages. In the evaluation, we collect a few hundred bloggers from a well-known Sina blog host bloggers categories in China, and an also well-known blogger community service Nihon Blog Mura in Japan. As case studies, we focus on the "health", "military", and "nursing care" categories in the services of both languages, and generate topics based on a topic model, and then overview and compare them between Chinese and Japanese. We actually discover certain differences in bloggers' topics between Chinese and Japanese.