POS Hindi English

Each token in the dataset is labeled with a language code like 'en' for English, 'hi' for Hindi, and 'rest' for other or unclassified tokens. Corresponding entity tags follow the BIO tagging format, denoting the beginning (B), inside (I), or outside (O) of named entities such as PERSON, PLACE, or ORGANISATION. The data is sourced from social platforms like Twitter and includes features like mentions, hashtags, emojis, and links, reflecting real-world language complexity.