Abstract
Structural divergence presents a challenge to the use of syntax in statistical machine translation. We address this problem with a new algorithm for alignment of loosely matched non-isomorphic dependency trees. The algorithm selectively relaxes the constraints of the two tree structures while keeping computational complexity polynomial in the length of the sentences. Experimentation with a large Chinese-English corpus shows an improvement in alignment results over the unstructured models of (Brown et al., 1993).- Anthology ID:
- 2003.mtsummit-papers.13
- Volume:
- Proceedings of Machine Translation Summit IX: Papers
- Month:
- September 23-27
- Year:
- 2003
- Address:
- New Orleans, USA
- Venue:
- MTSummit
- SIG:
- Publisher:
- Note:
- Pages:
- Language:
- URL:
- https://aclanthology.org/2003.mtsummit-papers.13
- DOI:
- Cite (ACL):
- Yuan Ding, Daniel Gildea, and Martha Palmer. 2003. An algorithm for word-level alignment of parallel dependency trees. In Proceedings of Machine Translation Summit IX: Papers, New Orleans, USA.
- Cite (Informal):
- An algorithm for word-level alignment of parallel dependency trees (Ding et al., MTSummit 2003)
- PDF:
- https://preview.aclanthology.org/remove-xml-comments/2003.mtsummit-papers.13.pdf