People are increasingly sharing their personal health information in online chats, blogs and social networks. To prevent a possible damage to individuals, we want to develop tools that can help to anonymize posts on such medical forums. In the first step of our study, we work on identification of the authors of the posts using statistical methods. In the next step we explore author’s writing style.

The questions we aim to answer are: (1) How well can the author of the short user-generated web post be identified by automated methods? (2) What text features can reveal author’s specific writing? (3) How can a text be anonymized?