Sentiment Analysis of Depression and Anxiety Social Media Tweets Using TF-IDF Weighting and Supervised Learning Algorithm

One of the most prevalent mental illnesses and a significant contributing factor to suicides is depression. The increased usage of social media platforms has made it possible to diagnose depression early on by speaking to individuals in their native tongue. This research explores sentiment analysis on a Twitter dataset by categorizing tweets into ‘positive’ and ‘negative’ sentiments. The proposed model exhibits exceptional performance in the context of depression and anxiety sentiment classification. Employing pre-processing techniques enhanced dataset quality, including tokenization, lemmatization, and stop word removal. The entire process, including the analysis and sorting steps, would use social media data, including text data and emoticons, emojis, and other graphics. Incorporating the TF-IDF weighting scheme further refined sentiment conveyance by emphasizing specific terms. A supervised machine-learning algorithm was applied, and the proposed model demonstrated superior performance, achieving precision, recall, F1-Score, and accuracy at 98.27%, 98.07%, 98.00%, and 98.01%, respectively. These remarkable metrics underscore the proposed model’s exceptional ability to classify tweets accurately, highlighting its robustness in handling X (formally Twitter) data distinctions for advanced social media analytics.