Corpus Creation and Annotating Multilingual Code-Mixed Kannada English Data with Precise Labels for Depression Detection
C H Shwetha, K. P. Pushpalatha · 2024
This study focuses on understanding mental health in multilingual communities. Specifically, it looks at how people make use of a multilingual language like Kannada and English while sharing their emotions and when talking about depression. The study creates a dataset with detailed labels to identify signs of depression and assess its severity. Through the examination of linguistic elements associated with depression, such as emotional expression and grammatical composition, this research contributes to a deeper understanding of how psychological wellbeing is conveyed across various languages. The labeled dataset plays a central role in developing deep learning algorithms capable of identifying and assessing depression. Additionally, the study emphasizes how language matters in mental health discussions and can influence clinical practices in diverse societies.