Skip to content

Latest commit

 

History

History
140 lines (96 loc) · 4.45 KB

README.md

File metadata and controls

140 lines (96 loc) · 4.45 KB
Raw

Build status Coverage status CPAN version

NAME

Lingua::Stem::UniNE - University of Neuchâtel stemmers

VERSION

This document describes Lingua::Stem::UniNE v0.08.

SYNOPSIS

use Lingua::Stem::UniNE;

# create Bulgarian stemmer
$stemmer = Lingua::Stem::UniNE->new(language => 'bg');

# get stem for word
$stem = $stemmer->stem($word);

# get list of stems for list of words
@stems = $stemmer->stem(@words);

DESCRIPTION

This module contains a collection of stemmers for multiple languages based on stemming algorithms provided by Jacques Savoy of the University of Neuchâtel (UniNE). The languages currently implemented are Bulgarian, Czech, German, and Persian. Work is ongoing for Arabic, Bengali, Finnish, French, Hindi, Hungarian, Italian, Portuguese, Marathi, Russian, Spanish, and Swedish. The top priority is languages for which there are no stemmers available on CPAN.

Attributes

  • language

    The following language codes are currently supported.

      ┌───────────┬────┐
  │ Bulgarian │ bg │
  │ Czech     │ cs │
  │ German    │ de │
  │ Persian   │ fa │
  └───────────┴────┘

    They are in the two-letter ISO 639-1 format and are case-insensitive but are always returned in lowercase when requested.

    # instantiate a stemmer object
$stemmer = Lingua::Stem::UniNE->new(language => $language);

# get current language
$language = $stemmer->language;

# change language
$stemmer->language($language);

    Country codes such as cz for the Czech Republic are not supported, nor are IETF language tags such as fa-AF or fa-IR.

  • aggressive

    By default, if there are multiple strengths of stemmers, a light stemmer will be used. When aggressive is set to true, an aggressive stemmer will be used if available.

    $stemmer->aggressive(1);

    Czech and German have aggressive options.

Methods

  • stem

    Accepts a list of words, stems each word, and returns a list of stems. The list returned will always have the same number of elements in the same order as the list provided. When no stemming rules apply to a word, the original word is returned.

    @stems = $stemmer->stem(@words);

# get the stem for a single word
$stem = $stemmer->stem($word);

    The words should be provided as character strings and the stems are returned as character strings. Byte strings in arbitrary character encodings are intentionally not supported.

  • languages

    Returns a list of supported two-letter language codes using lowercase letters.

    # object method
@languages = $stemmer->languages;

# class method
@languages = Lingua::Stem::UniNE->languages;

SEE ALSO

Lingua::Stem::Any provides a unified interface to any stemmer on CPAN, including this module, as well as additional features like normalization, casefolding, and in-place stemming.

Lingua::Stem::Snowball provides alternate stemming algorithms for Finnish, French, German, Hungarian, Italian, Portuguese, Russian, Spanish, and Swedish, as well as other languages.

These stemming algorithms are based on definition and implementations by Jacques Savoy and Ljiljana Dolamic of the University of Neuchâtel and provided at IR Multilingual Resources at UniNE.

AUTHOR

Nick Patch [email protected]

This project is brought to you by Shutterstock. Additional open source projects from Shutterstock can be found at code.shutterstock.com.

COPYRIGHT AND LICENSE

© 2012–2014 Shutterstock, Inc.

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.