morphology

morphology copied to clipboard

• Morphological Analyzer for Latvian Language *

A Java library for analyzing morphology and part of speech information for Latvian words. Accurate analysis is based on an included lexicon of some 50.000 lexemes, for rarer words there is some ambiguity. Also includes generation of all inflections of a word, and crude statistical disambiguation for analysis.

BASIC USAGE

Analyzer analyzer = new Analyzer();

// analysis Word result = analyzer.analyze("roku"); for (Wordform wf : result.wordforms) { wf.describe(); }

// generation of inflections List<Wordform> wordforms = analyzer.generateInflections("rakt"); for (Wordform wf : wordforms) { wf.describe(); }

Review unit tests for more examples.

INSTALLATION

Use maven to build and deploy.

LICENCE

(c) Institute of Mathematics and Computer Science, University of Latvia, 2005-2012

This software is licenced under GNU General Public Licence. Commercial licencing is available if neccessary, contact us at [email protected].

REFERENCES Current usage is described at http://www.ep.liu.se/ecp_article/index.en.aspx?issue=085;article=024 The initial core algorithm is published at http://www.semti-kamols.lv/doc_upl/Kamols-Kaunas-paper-3.pdf