© 2017 IEEE. In Japan, when a disaster occurs, victims and local governments use social network services (SNSs), such as twitter, as a means of information transmission. Therefore, a variety of information gathers on SNSs. It is considered that, by utilizing this information, assistance to victims can be offered effectively. In this study, we focused on the twitter SNS, and we propose a method of extracting information by utilizing correspondence analysis and its visualization. Furthermore, using this method, we developed a web application for extracting search effective information.