In recent years, people who are looking for a housing property by using the real estate portal site is rapidly increasing. On the other hand, it is difficult to read the background and needs of users from the site of the access log. Many users are writing to social media frequently that sometimes feelings and experiences. Therefore, by analyzing the write social media user using real estate portal site, we considered more appropriate recommendation of housing property can be achieved. In this study, the use of crowdsourcing from writing information of social media in order to make an estimate of the housing property search phase. As a result, it was confirmed that discover the user's housing property background and needs.
