Bilingual multi-word lexicon construction via a pivot language

Abstract

Bilingual multi-word lexicons are helpful for statistical machine translation systems to improve their performance. In this paper we present a method for constructing such lexicons in a resource-poor language pair such as Korean-French. By using two parallel corpora sharing one pivot language we can easily construct such lexicons without any external language resource like a seed dictionary. The experimental results for the KR to FR have shown that the accuracy is quite promising, even though this research is ongoing.

Bilingual Multiword Expression Alignment by Constituent-Based Similarity Score

“…Based on these circumstances, we identified potential MWE candidates for source and target languages by using both linguistic and statistical information [19,20]. The identification method (see [11] for more details) can be described as follows:…”

“…This is one of the disadvantages for extracting bilingual MWEs from parallel corpora. Under the circumstance, Seo et al [11] have proposed a novel method, denoted as the pivot context-based approach for multiwords (PCAM) in the rest of the paper, for extracting bilingual MWEs by using parallel corpora in a resource-poor language pair. However, PCAM has a weak point that targets constituents instead of complete translation equivalents, which might have high scores of similarity if their contexts are not enough.…”