2026-07-22T12:14:55Z https://drops.dagstuhl.de/oai

oai:drops-oai.dagstuhl.de:4044 2024-03-06T10:29:12Z ddc:004 open_access

Combining Language Independent Part-of-Speech Tagging Tools Orosz, György Laki, László János Novák, Attila Siklósi, Borbála part-of-speech tagging combination agglutinative languages machine learning machine translation Part-of-speech tagging is a fundamental task of natural language processing. For languages with a very rich agglutinating morphology, generic PoS tagging algorithms do not yield very high accuracy due to data sparseness issues. Though integrating a morphological analyzer can efficiently solve this problem, this is a resource-intensive solution. In this paper we show a method of combining language independent statistical solutions -- including a statistical machine translation tool -- of PoS-tagging to effectively boost tagging accuracy. Our experiments show that, using the same training set, our combination of language independent tools yield an accuracy that approaches that of a language dependent system with an integrated morphological analyzer. Schloss Dagstuhl – Leibniz-Zentrum für Informatik György Orosz and László János Laki and Attila Novák and Borbála Siklósi 2013 Is Part Of OASIcs, Volume 29, 2nd Symposium on Languages, Applications and Technologies (2013) InProceedings Text doc-type:ResearchArticle publishedVersion application/pdf doi:10.4230/OASIcs.SLATE.2013.249 urn:nbn:de:0030-drops-40441 https://drops.dagstuhl.de/entities/document/10.4230/OASIcs.SLATE.2013.249 eng https://creativecommons.org/licenses/by/3.0/legalcode