Analyse des Tweets pour la détection des troubles mentaux (dépression et anxiété)
No Thumbnail Available
Date
2023
Journal Title
Journal ISSN
Volume Title
Publisher
University of Oum El Bouaghi
Abstract
Social media occupies an important placein daily life, where users share different elements such as thoughts, experiences, events and feelings. The massive use of social media has led to the generation of huge volumes of data. These data constitute a treasure allowing the extraction of relevant information through natural language processing tasks, in particular by involving deep learning techniques. Starting from this context, various researches have been carried out with the aim of studying the detection of mental disorders, especially depression and anxiety, through the analysis of data extracted from Twitter platform. Although these researches have been able to achieve very satisfactory results, they nevertheless rely on binary classification models by treating each mental disorder separately. Indeed, it would be better if we manage to develop systems capable of dealing with several mental disorders. To address this point, we propose a well-defined methodology involving the use of deep learning to develop effective multi-class modelsfor detecting depression and anxiety through the analysis oftweets. The idea consists in testingboth simple and hybrid variants of deep learning models to examine their strengths and weaknesses. Once the models are built, we move on to the deployment phase in which the best performing models are used to analyze tweets in the UK and US during the COVID-19 period regarding depressive and anxious behaviors. Our work is validated by several experiments and comparisons. Globally, the results obtained are satisfactory and encouraging