Standard sentiment analysis in Indian vernaculars relies on flat, one-dimensional classification. The Maithili Multi-Dimensional Sentiment Corpus (64,215 rows) introduces a high-resolution, socio-linguistically grounded architecture. It utilizes a dual-axis classification system—predicting both primary sentiment and emotional intensity—mapped across highly… See the full description on the dataset page: https://huggingface.co/datasets/abhiprd2000/Maithili_Sentiment_8K.
