 | |
| Original author(s) | Matthew Honnibal |
|---|---|
| Developer(s) | Explosion AI, various |
| Initial release | February 2015; 10 years ago (2015-02)[1] |
| Stable release | |
| Repository | |
| Written in | Python,Cython |
| Operating system | Linux,Windows,macOS,OS X |
| Platform | Cross-platform |
| Type | Natural language processing |
| License | MIT License |
| Website | spacy |
spaCy (/speɪˈsiː/spay-SEE) is anopen-source software library for advancednatural language processing, written in the programming languagesPython andCython.[3][4] The library is published under theMIT license and its main developers areMatthew Honnibal andInes Montani, the founders of the software company Explosion.
UnlikeNLTK, which is widely used for teaching and research, spaCy focuses on providing software for production usage.[5][6] spaCy also supportsdeep learning workflows that allow connecting statistical models trained by popularmachine learning libraries likeTensorFlow,PyTorch orMXNet through its own machine learning library Thinc.[7][8] Using Thinc as its backend, spaCy featuresconvolutional neural network models forpart-of-speech tagging,dependency parsing,text categorization andnamed entity recognition (NER). Prebuilt statisticalneural network models to perform these tasks are available for 23 languages, including English, Portuguese, Spanish, Russian and Chinese, and there is also a multi-languageNER model. Additional support fortokenization for more than 65 languages allows users to train custom models on their own datasets as well.[9]
spaCy comes with several extensions and visualizations that are available as free,open-source libraries: