LIMA - Libre Multilingual Analyzer

LIMA is a multilingual natural language processing analyzer developed by CEA LIST (LASTI laboratory). It is free software, available under the MIT license.

• Neural modules (libtorch) for tokenization, part-of-speech tagging and morphological features, lemmatization and dependency parsing, with models for more than 60 languages.
• ModEx, a powerful rule-based mechanism to extract entities, relations and events in domains where no annotated data exists.

Documentation

https://aymara.github.io/lima/: installation, user guide, guides, reference and developer documentation.

Quick start

With the Python package (Linux x86_64, Python ≥ 3.7):

pip install --upgrade pip
pip install aymara==0.5.0b6
lima_models -i eng

import aymara.lima
nlp = aymara.lima.Lima("ud-eng")
doc = nlp("Hello, World!")
print(repr(doc))

Other options (Docker image, building from source) are described in the installation guide.

Contributing

Bug reports and contributions are welcome: see the contributing guide. The documentation sources are in docs/.