Knowledge-based machine translation (KBMT) techniques yield high quality in domains with detailed semantic models, limited vocabulary, and controlled input grammar. Scaling up along these dimensions means acquiring large knowledge resources. It also means behaving reasonably when definitive knowledge is not yet available. This paper describes how we can fill various KBMT knowledge gaps, often using robust statistical techniques. We describe quantitative and qualitative results from JAPANGLOSS, a broad-coverage Japanese-English MT system.
@article{arxiv.cmp-lg/9506009,
title = {Filling Knowledge Gaps in a Broad-Coverage Machine Translation System},
author = {Kevin Knight and Ishwar Chander and Matthew Haines and Vasileios Hatzivassiloglou and Eduard Hovy and Masayo Iida and Steve K. Luk and Richard Whitney and Kenji Yamada},
journal= {arXiv preprint arXiv:cmp-lg/9506009},
year = {2008}
}
Comments
7 pages, Compressed and uuencoded postscript. To appear: IJCAI-95