An extremely weakly-supervised text classification method using mutually-enhancing text granularities (word, sentence, and document-level context).
Primary LanguagePython