Experiments in cross-language morphological annotation transferz

Anna Feldman, Jirka Hana, Chris Brew

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations


Annotated corpora are valuable resources for NLP which are often costly to create. We introduce a method for transferring annotation from a morphologically annotated corpus of a source language to a target language. Our approach assumes only that an unannotated text corpus exists for the target language and a simple textbook which describes the basic morphological properties of that language is available. Our paper describes experiments with Polish, Czech, and Russian. However, the method is not tied in any way to these languages. In all the experiments we use the TnT tagger ([3]), a second-order Markov model. Our approach assumes that the information acquired about one language can be used for processing a related language. We have found out that even breath-takingly naive things (such as approximating the Russian transitions by Czech and/or Polish and approximating the Russian emissions by (manually/automatically derived) Czech cognates) can lead to a significant improvement of the tagger's performance.

Original languageEnglish
Title of host publicationComputational Linguistics and Intelligent Text Processing - 7th International Conference, CICLing 2006, Proceedings
Number of pages10
StatePublished - 2006
Event7th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2006 - Mexico City, Mexico
Duration: 19 Feb 200625 Feb 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3878 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other7th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2006
CityMexico City


Dive into the research topics of 'Experiments in cross-language morphological annotation transferz'. Together they form a unique fingerprint.

Cite this