Sentiment, Emotion analysis for mental health based on text

This dataset contains 160,000 social media text records curated and customized from two distinct sources to facilitate advanced emotion recognition and sentiment analysis tasks. The text entries capture nuanced human emotions and psychological states expressed on digital platforms. Each record is categorized into one of ten specific classes, spanning standard emotions, general sentiments, and critical psychological states like depression. This curated dataset is highly suitable for training machine learning models, deep learning architectures (e.g., BERT, RoBERTa, BiLSTM), and ensemble meta-classifiers.