Your browser doesn't support javascript.
Comparative Analysis of Deep-Learning techniques for Depressive Text Classification
7th IEEE International conference for Convergence in Technology, I2CT 2022 ; 2022.
Article in English | Scopus | ID: covidwho-1992598
ABSTRACT
Psychiatric problems and disorders are an epidemic in their own right, however, they often go unnoticed & undetected. Post COVID-19, given that most families & individuals were forced to isolate themselves in their homes for huge periods of time, it only made things worse. We humans being social animals needed a refuge, therefore the volume of interactions & personal posts on social media platforms skyrocketed. Whilst, there's huge leverage of text classification techniques using Deep Learning algorithms for financial, commercial applications eg. stock market news analysis, analysing customer behavior, etc. but similar applications in the field of Mental health are quite meager. The text on the social media feed of a person can be critical and of huge help in expeditious detection of depressive disorders. Via the medium of this paper, our aim is to find an optimum solution for the above-addressed problem, we take the real-time user data from an online social networking platform, after which it is pre-processed and analysed, thereafter we use this data to build deep learning-based classifier models i.e. LSTM, (CNN+LSTM), GRU, these models are improved using optimisation algorithms, furthermore, these models are compared and analysed to check which text classification algorithm works best for our use case. © 2022 IEEE.
Keywords

Full text: Available Collection: Databases of international organizations Database: Scopus Language: English Journal: 7th IEEE International conference for Convergence in Technology, I2CT 2022 Year: 2022 Document Type: Article

Similar

MEDLINE

...
LILACS

LIS


Full text: Available Collection: Databases of international organizations Database: Scopus Language: English Journal: 7th IEEE International conference for Convergence in Technology, I2CT 2022 Year: 2022 Document Type: Article