spaCy

{{Short description|Software library for natural language processing}}

{{lowercase title}}

{{For|the 1981 film|Spacy (film) {{!}} Spacy (film)}}

{{Distinguish|Scapy}}

{{Infobox software

| title = spaCy

| name = spaCy

| logo = SpaCy logo.svg

| author = Matthew Honnibal

| developer = Explosion AI, various

| released = {{Start date and age|2015|02}}{{cite web |title=Introducing spaCy |publisher=explosion.ai |url=https://explosion.ai/blog/introducing-spacy|access-date=2016-12-18}}

| latest release version = {{wikidata|property|reference|edit| Q28406945 |P348}}

| latest release date = {{start date and age|{{wikidata|qualifier| Q28406945 |P348|P577}}}}

| programming language = Python, Cython

| operating system = Linux, Windows, macOS, OS X

| platform = Cross-platform

| genre = Natural language processing

| license = MIT License

| website = {{Official URL}}

}}

spaCy ({{IPAc-en|s|p|eɪ|ˈ|s|iː}} {{respell|spay|SEE|'}}) is an open-source software library for advanced natural language processing, written in the programming languages Python and Cython.Choi et al. (2015). [https://aclweb.org/anthology/P/P15/P15-1038.pdf It Depends: Dependency Parser Comparison Using A Web-based Evaluation Tool].{{Cite web|url=https://www.washingtonpost.com/news/wonk/wp/2016/05/18/googles-new-artificial-intelligence-cant-understand-these-sentences-can-you/|title=Google's new artificial intelligence can't understand these sentences. Can you?|website=Washington Post|access-date=2016-12-18}} The library is published under the MIT license and its main developers are Matthew Honnibal and Ines Montani, the founders of the software company Explosion.

Unlike NLTK, which is widely used for teaching and research, spaCy focuses on providing software for production usage.{{Cite web|url=https://spacy.io/usage/facts-figures#other-libraries|title=Facts & Figures - spaCy|website=spacy.io|language=en|access-date=2020-04-04}}{{cite journal|last=Bird|first=Steven|author2=Klein, Ewan|author3=Loper, Edward|author4=Baldridge, Jason|year=2008|title=Multidisciplinary instruction with the Natural Language Toolkit|url=https://www.aclweb.org/anthology/W/W08/W08-0208.pdf|journal=Proceedings of the Third Workshop on Issues in Teaching Computational Linguistics, ACL|page=62|doi=10.3115/1627306.1627317|isbn=9781932432145|s2cid=16932735|doi-access=free}} spaCy also supports deep learning workflows that allow connecting statistical models trained by popular machine learning libraries like TensorFlow, PyTorch or MXNet through its own machine learning library Thinc.{{Cite web|url=https://thinc.ai/docs/usage-frameworks|title=PyTorch, TensorFlow & MXNet|website=thinc.ai|access-date=2020-04-04}}{{Cite web|url=https://github.com/explosion/thinc|title=explosion/thinc|website=GitHub|access-date=2016-12-30}} Using Thinc as its backend, spaCy features convolutional neural network models for part-of-speech tagging, dependency parsing, text categorization and named entity recognition (NER). Prebuilt statistical neural network models to perform these tasks are available for 23 languages, including English, Portuguese, Spanish, Russian and Chinese, and there is also a multi-language NER model. Additional support for tokenization for more than 65 languages allows users to train custom models on their own datasets as well.{{Cite web|url=https://spacy.io/usage/models#languages|title=Models & Languages {{!}} spaCy Usage Documentation|website=spacy.io|access-date=2020-03-10}}