Combining bilingual terminology mining and morphological modeling for domain adaptation in SMT
Marion Weller, Alexander Fraser, Ulrich Heid · 2014
Translating in technical domains is a well-known problem in SMT, as the lack of par-allel documents causes significant problems of sparsity. We discuss and compare differ-ent strategies for enriching SMT systems built on general domain data with bilingual terminology mined from comparable cor-pora. In particular, we focus on the target-language inflection of the terminology data and present a pipeline that can generate pre-viously unseen inflected forms. 1