Discovery of Informal Topics from Post Traumatic Stress Disorder Forums
Discovery of Informal Topics from Post Traumatic Stress Disorder Forums
复制标题
DOI：
10.1109/icdmw.2017.65
复制
发表时间：
2017-11
期刊：
影响因子：
--
通讯作者：
Reilly Grant;David Kucher;A. Leon;Jonathan F. Gemmell;D. Raicu
中科院分区：
文献类型：
--
作者：
Reilly Grant;David Kucher;A. Leon;Jonathan F. Gemmell;D. Raicu
创伤后应激障碍（PTSD）是一种公共卫生问题，每年困扰着数百万人。这在退伍军人中尤为突出。理解与PTSD相关的语言，态度和主题是一个重要而具有挑战性的问题。根据他们的专业知识，心理健康专家已经构建了PTSD的正式定义。然而，即使是最勤奋的心理健康专业人员也只能照顾一小部分患有PTSD的人，限制了他们对这种疾病的看法。随着社交网站越来越被接受，用户开始表达个人的想法和感受，比如那些与创伤后应激障碍有关的想法和感受。这些丰富的内容可以被视为对创伤后应激障碍及其相关问题的大量集体描述。我们自动从成千上万的社交媒体帖子中提取非正式的潜在主题，其中用户描述了他们的PTSD经历，并将这些主题与心理健康专业人员生成的正式描述进行比较。然后，我们探讨这些主题的模式和关联。我们的非正式主题发现评估表明，我们可以成功地识别PTSD社交媒体相关数据中有意义的主题。当将我们的主题与《精神疾病诊断与统计手册》（DSM）中的标准进行比较时，我们发现我们能够自动复制许多标准。我们还发现了DSM中没有提到的新主题，但在数千名用户的PTSD体验的协作叙述中普遍存在。
Post Traumatic Stress Disorder (PTSD) is a public health problem afflicting millions of people each year. It is especially prominent among military veterans. Understanding the language, attitudes, and topics associated with PTSD presents an important and challenging problem. Based on their expertise, mental health professionals have constructed a formal definition of PTSD. However, even the most assiduous mental health professionals can care for only a small fraction of those suffering from PTSD, limiting their perspective of the disorder. As social networking sites have grown in acceptance, users have begun to express personal thoughts and feelings, such as those related to PTSD. This wealth of content can be viewed as an enormous collective description of PTSD and its related issues. We automatically extract informal latent topics from thousands of social media posts in which users describe their experience with PTSD and compare these topics to the formal description generated by mental health professionals. We then explore the pattern and associations of these topics. Our informal topic discovery evaluation reveals that we can successfully identify meaningful topics in PTSD social media related data. When comparing our topics to the criteria included in the Diagnostic and Statistical Manual of Mental Disorders (DSM), we found that we were able to automatically reproduce many of the criteria. We also discovered new topics which were not mentioned in the DSM, but were prevalent across the collaborative narrative of thousands of user's experience with PTSD.