Detecting candidates of depression, anxiety and stress through malay-written tweets: a preliminary study

Abstract

<p>Depression, anxiety and stress are not trivial conditions applicable for only the weak-hearted. They can be inflicted by anyone of all age groups, gender, race and social status. While some are courageous to acknowledge their condition, others shy away in shame or denial. In this paper, we proposed a “proactive” approach to detecting candidates of depression, anxiety and stress in an unobtrusive manner by tapping into what Malaysians tweet in Malay language. From this preliminary study, we constructed 165 Malay layman terms which describe depression, anxiety or stress as identified in M-DASS-42 scale. Since Twitter is an informal platform, construction of Malay layman terms is an essential step to the detection of candidates. Our study on 1,789 Malay tweets discovered 6 Twitter users as potential candidates, having high frequency of tweets with any of the layman terms. We can conclude that using tweets can be useful in unobtrusively detecting candidates of depression, anxiety or stress. This paper also identifies open research areas.</p>

Stochastic modelling of transition dynamic of mixed mood episodes in bipolar disorder

“…The best representation can be given by ICD-10 when there are two significant episodes of mood where manic episodes are mandatory to be included in order to diagnose bipolar disorder [4]. There are different forms of clinical biomarkers which are used in diagnosis process where neuroimaging contributes to understand the distinction between unipolar and bipolar depression [5]. However, it is strongly felt that this distinction is never clear enough for understanding the complexities of mood transition in bipolar disorder.…”