Mood disorder patients' language features on their microblogs
Tai Wang, Zongkui Zhou, Tingshao Zhu, Wei Yang · International Journal of Embedded Systems · 2014
People's language features are exhibited on their online social network websites, such as Twitter, Weibo in Sina or ShuoShuo in QQ (a former version of microblog). Several leading labs have already made remarkable breakthroughs in the area of collecting and analysing texts generated by a huge population. In this paper, a novel research topic is presented, with the assumption that different kinds of people may exhibit their unique language features, especially mood disorder patients and normal people. The best efforts have been carried out to verify this assumption. Three mood disorder patients and 32 normal people are invited into this test, with their four–year short texts on their microblogs. The results show that though there is no obvious difference between their neither positive nor negative emotion expressions, a sharp gap does exist in the dimension of anger. The authors expect their findings can be tested in a much larger dataset in the future. If the conclusion still holds, a promising auxiliary toolkit for mood disorder diagnosis can thus be developed.