stanza

Why we included this project

Stanza is the Stanford NLP Group's official Python library, a dependable way to run a complete linguistic pipeline without assembling your own tokenizer and tagger stack. Point it at raw text in any of 60+ languages and it will split sentences, segment tokens, label parts of speech, recognize named entities, and produce dependency parses, all from a single pip install. Developers and small teams doing multilingual preprocessing for search, extraction, or corpus work get output that follows well-understood conventions, since the models train on Universal Dependencies and other standardized resources. It also wraps the Java Stanford CoreNLP tools, which helps when you need to bridge Python code with existing CoreNLP annotations. The newer biomedical and clinical English model packages bring the same interface to a domain where accurate parsing and entity recognition on research papers and clinical notes matters a great deal.