Identifying Privacy Leakage from User-Generated Content in an Online Health Community

Abstract

Online Health Communities (OHCs) have become a widely used resource for obtaining and sharing health-related information during the past decade. However, the health information privacy issues of OHCs have not been fully explored. Insufficient attention to personal privacy management may result in intentional or unintentional disclosure of users’ sensitive information, and consequently harm the communication environment, as well as individuals’ safety. Based on the user-generated content, this preliminary research applies the method of text mining to identify different types of information leakages occur in a breast cancer OHC. The results indicate that approximately 60% of the OHC users are willing to express their emotional feelings, and 10.86% are motivated to disclose their health information. In addition, based on the longitudinal data from 2007 to 2018, we analyzed the OHC user behavior trajectories in private information exposure. The findings of this study have practical implications for OHC users, administers, and website designers.