Major depressive disorder is one of the most crippling diseases, it accounts for 4.3% of the global disease burden. With depression being so prevalent, it is of paramount importance that a systematic method of diagnosis should exist. However, current diagnostic methods such as questionnaires and clinical diagnosis, consist of self-reported symptoms. Therefore, both methods are susceptible to patient manipulation. Nowadays social media websites such as Facebook, Twitter, Reddit, and Tumblr provide a method to obtain behavioral attributes to thoughts and interactions of a person. This research is focused on developing a machine learning model capable of analyzing linguistic patterns obtained from Twitter user data and determining whether a particular user exhibits depressive symptoms. We trained Support Vector Machine and Random Forest models for this purpose and, compare their efficiency of diagnosis and concluded that Random Forest gave the best results. We believe the results obtained from this research can be considered for the development of the new technique for effective identification of depressed users on social media platforms.
Paper
Full text
Identifying Depression Among Twitter Users using Sentiment Analysis
Semantic Scholar · Computer Science · 2021
Abstract
Major depressive disorder is one of the most crippling diseases, it accounts for 4.3% of the global disease burden. With depression being so prevalent, it is of paramount importance that a systematic method of diagnosis should exist. However, current diagnostic methods such as questionnaires and clinical diagnosis, consist of self-reported symptoms. Therefore, both methods are susceptible to patient manipulation. Nowadays social media websites such as Facebook, Twitter, Reddit, and Tumblr provide a method to obtain behavioral attributes to thoughts and interactions of a person. This research is focused on developing a machine learning model capable of analyzing linguistic patterns obtained from Twitter user data and determining whether a particular user exhibits depressive symptoms. We trained Support Vector Machine and Random Forest models for this purpose and, compare their efficiency of diagnosis and concluded that Random Forest gave the best results. We believe the results obtained from this research can be considered for the development of the new technique for effective identification of depressed users on social media platforms.