Skip to main content

Stripped down, forked version of Pattern package (Web mining module for Python.)

Project description

PatternLite

PyPi version License

This is a fork of the original Pattern package. It is a stripped down version of the original package and contains only the text modules (i.e. pattern.de, pattern.en, pattern.es, pattern.fr, pattern.it, pattern.nl, pattern.ru, the vector module and parts of the db module.

Apart from that, it fixes a few problems with the original package:

  • much fewer dependencies: now only requires numpy, scipy and nltk
  • compatible with Python 3.6, 3.7, 3.8
  • properly closes files (no more "ResourceWarning: unclosed file")

The remaining original features are:

  • Natural Language Processing: part-of-speech taggers, n-gram search, sentiment analysis, WordNet
  • Machine Learning: vector space model, clustering, classification (KNN, SVM, Perceptron)

For the original documentation, see http://www.clips.ua.ac.be/pages/pattern.

Installation

PatternLite supports Python 3.6, 3.7 and 3.8.

If you have pip, you can automatically download and install from the PyPI repository:

pip install PatternLite

To install PatternLite from source, unzip the download and from the command line do:

python setup.py install

Documentation

For documentation and examples see the user documentation. If you are a developer, go check out the developer documentation.

Version

3.7

License

BSD, see LICENSE.txt for further details.

Reference

De Smedt, T., Daelemans, W. (2012). Pattern for Python. Journal of Machine Learning Research, 13, 2031–2035.

Contribute

The source code is hosted on GitHub and contributions or donations are welcomed. Please have look at the developer documentation. If you use Pattern in your work, please cite our reference paper.

Bundled dependencies

Pattern is bundled with the following data sets, algorithms and Python packages:

  • Brill tagger, Eric Brill
  • Brill tagger for Dutch, Jeroen Geertzen
  • Brill tagger for German, Gerold Schneider & Martin Volk
  • Brill tagger for Spanish, trained on Wikicorpus (Samuel Reese & Gemma Boleda et al.)
  • Brill tagger for French, trained on Lefff (Benoît Sagot & Lionel Clément et al.)
  • Brill tagger for Italian, mined from Wiktionary
  • English pluralization, Damian Conway
  • Spanish verb inflection, Fred Jehle
  • French verb inflection, Bob Salita
  • LIBSVM, Chih-Chung Chang & Chih-Jen Lin
  • LIBLINEAR, Rong-En Fan et al.
  • spelling corrector, Peter Norvig

Acknowledgements

Authors:

Contributors (chronological):

  • Frederik De Bleser
  • Jason Wiener
  • Daniel Friesen
  • Jeroen Geertzen
  • Thomas Crombez
  • Ken Williams
  • Peteris Erins
  • Rajesh Nair
  • F. De Smedt
  • Radim Řehůřek
  • Tom Loredo
  • John DeBovis
  • Thomas Sileo
  • Gerold Schneider
  • Martin Volk
  • Samuel Joseph
  • Shubhanshu Mishra
  • Robert Elwell
  • Fred Jehle
  • Antoine Mazières + fabelier.org
  • Rémi de Zoeten + closealert.nl
  • Kenneth Koch
  • Jens Grivolla
  • Fabio Marfia
  • Steven Loria
  • Colin Molter + tevizz.com
  • Peter Bull
  • Maurizio Sambati
  • Dan Fu
  • Salvatore Di Dio
  • Vincent Van Asch
  • Frederik Elwert

Release history Release notifications | RSS feed

This version

3.6

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

PatternLite-3.6.tar.gz (22.0 MB view details)

Uploaded Source

Built Distribution

PatternLite-3.6-py3-none-any.whl (22.1 MB view details)

Uploaded Python 3

File details

Details for the file PatternLite-3.6.tar.gz.

File metadata

  • Download URL: PatternLite-3.6.tar.gz
  • Upload date:
  • Size: 22.0 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.1.1 pkginfo/1.5.0.1 requests/2.22.0 setuptools/42.0.2 requests-toolbelt/0.9.1 tqdm/4.40.0 CPython/3.6.9

File hashes

Hashes for PatternLite-3.6.tar.gz
Algorithm Hash digest
SHA256 bfc890760dd2dfec89f0b01167ef9a460b31ecb9ad8cebcd4b3798a96a62e260
MD5 09369eb8414477a6cb01afae24c9ff33
BLAKE2b-256 5a65d872a6269072568f4e812fc071705b5134c6153e05b121ed3234a16eb1ca

See more details on using hashes here.

File details

Details for the file PatternLite-3.6-py3-none-any.whl.

File metadata

  • Download URL: PatternLite-3.6-py3-none-any.whl
  • Upload date:
  • Size: 22.1 MB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.1.1 pkginfo/1.5.0.1 requests/2.22.0 setuptools/42.0.2 requests-toolbelt/0.9.1 tqdm/4.40.0 CPython/3.6.9

File hashes

Hashes for PatternLite-3.6-py3-none-any.whl
Algorithm Hash digest
SHA256 c54cd29314635b0e6361861c33f036705c5449f08741c47d834317ba83bd98e0
MD5 73eda409bfca5726f817f8684f6cf136
BLAKE2b-256 50f51894eb24102cae0e433c18366ec2a8d945b42cbf128303b67454db8587d8

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page