Difference between revisions of "Resources for Polish"

From ACL Wiki

Revision as of 11:19, 12 October 2013

Europarl corpus, sentence aligned with English
IPI PAN Corpus - The IPI PAN Corpus is a large (currently over 250 million segments), morphosyntactically annotated, publicly available corpus of Polish, developed by the Linguistic Engineering Group at the Institute of Computer Science, Polish Academy of Sciences (ICS PAS)
PWN Corpus - PWN has prepared and made available an online version of the Corpus of Polish consisting of 40 million words. The samples were taken from 386 books, 977 editions selected from 185 different press publications, 84 transcribed spoken texts, 207 web sites and several hundred advertising leaflets and other ephemera. The full version of the corpus is available on payment for access, while a demonstration version of over 7.5 million words is available free of charge.

Morfologik -- morphological dictionary by Marcin Miłkowski (of LanguageTool), licensed under CC-SA / GNU LGPL
- Morfologik converted to the IKIPI tagset (the tagset of the IPI PAN Corpus)
Morphosyntactic Toolchain by WrocUT Language Technology Group G4.19, licensed under GNU LGPL (some optional addons are GNU GPL). Command-line utilities providing tokenisation, morphological analysis, morphosyntactic tagging, shallow parsing (chunking), WCCL feature vectors for machine learning.

"Morfeusz" - morphological analyser of Polish (Wolinski, 2005),
- main reference Morfeusz — a Practical Tool for the Morphological Analysis of Polish
"AMOR" - morphology analyser of Polish (Joanna Rabiega, 2000),
- main reference Podstawy lingwistyczne automatycznego analizatora morfologicznego AMOR
"SAM" - morphological analyser of Polish (Krzysztof Szafran, 1994),
Morfologik - Polish morphological analyzer based on current ispell dictionaries, and Java libraries interfacing it. First completely open-source and comprehensive morphological tools for Polish. Will be used for grammar correction tools (to be included in the future)
Spejd - Shallow Parsing and Disambiguation Engine
lemmatizer - Dawid Weiss

@@ Line 1: / Line 1: @@
 ==Corpora==
+* [http://www.statmt.org/europarl Europarl corpus], sentence aligned with English
 * [http://korpus.pl/en/ IPI PAN Corpus] - The IPI PAN Corpus is a large (currently over 250 million segments), morphosyntactically annotated, publicly available corpus of Polish, developed by the Linguistic Engineering Group at the  Institute of Computer Science, Polish Academy of Sciences (ICS PAS)
 * [http://korpus.pwn.pl/index_en.php PWN Corpus] - PWN has prepared and made available an online version of the Corpus of Polish consisting of 40 million words. The samples were taken from 386 books, 977 editions selected from 185 different press publications, 84 transcribed spoken texts, 207 web sites and several hundred advertising leaflets and other ephemera. The full version of the corpus is available on payment for access, while a demonstration version of over 7.5 million words is available free of charge.