Identifying word correspondence in parallel texts
HLT '91 Proceedings of the workshop on Speech and Natural Language
A systematic comparison of various statistical alignment models
Computational Linguistics
The mathematics of statistical machine translation: parameter estimation
Computational Linguistics - Special issue on using large corpora: II
Flow network models for word alignment and terminology extraction from bilingual corpora
COLING '98 Proceedings of the 17th international conference on Computational linguistics - Volume 1
Towards automatic extraction of monolingual and bilingual terminology
COLING '94 Proceedings of the 15th conference on Computational linguistics - Volume 1
Phrasal cohesion and statistical machine translation
EMNLP '02 Proceedings of the ACL-02 conference on Empirical methods in natural language processing - Volume 10
An evaluation exercise for word alignment
HLT-NAACL-PARALLEL '03 Proceedings of the HLT-NAACL 2003 Workshop on Building and using parallel texts: data driven machine translation and beyond - Volume 3
ProAlign: shared task system description
HLT-NAACL-PARALLEL '03 Proceedings of the HLT-NAACL 2003 Workshop on Building and using parallel texts: data driven machine translation and beyond - Volume 3
Using bilingual dependencies to align words in Enlish/French parallel corpora
ACLstudent '05 Proceedings of the ACL Student Research Workshop
Multi-task learning for word alignment and dependency parsing
AICI'11 Proceedings of the Third international conference on Artificial intelligence and computational intelligence - Volume Part III
| Hi-index | 0.00 |
We present a word alignment procedure based on a syntactic dependency analysis of French/English parallel corpora called "alignment by syntactic propagation". Both corpora are analysed with a deep and robust parser. Starting with an anchor pair consisting of two words which are potential translations of one another within aligned sentences, the alignment link is propagated to the syntactically connected words. The method was tested on two corpora and achieved a precision of 94.3 and 93.1% as well as a recall of 58 and 56%, respectively for each corpus.