Learning Phrase Translation using Level of Detail Approach
Hendra Setiawan, Haizhou Li, Min Zhang · 2005
We propose a simplified Level Of Detail (LOD) algorithm to learn phrase translation for statis-tical machine translation. In particular, LOD learns unknown phrase translations from par-allel texts without linguistic knowledge. LOD uses an agglomerative method to attack the combinatorial explosion that results when gen-erating candidate phrase translations. Al-though LOD was previously proposed by (Se-tiawan et al., 2005), we improve the original al-gorithm in two ways: simplifying the algorithm and using a simpler translation model. Experi-mental results show that our algorithm provides comparable performance while demonstrating a significant reduction in computation time. 1