Cross-language Projection of Dependency Trees with Constrained Partial Parsing for Tree-to-Tree Machine Translation

Abstract

Tree-to-tree machine translation (MT) that utilizes syntactic parse trees on both source and target sides suffers from the non-isomorphism of the parse trees due to parsing errors and the difference of annotation criterion between the two languages. In this paper, we present a method that projects dependency parse trees from the language side that has a high quality parser, to the side that has a low quality parser, to improve the isomorphism of the parse trees. We first project a part of the dependencies with high confidence to make a partial parse tree, and then complement the remaining dependencies with partial parsing constrained by the already projected dependencies. MT experiments verify the effectiveness of our proposed method.

Constrained Partial Parsing Based Dependency Tree Projection for Tree-to-Tree Machine Translation

“…We observe significant MT performance improvements on both the Japanese-Chinese and English-Chinese translation tasks in our large scale experiments, indicating the language and task independency of our proposed method. In addition, compared to our previous work (Shen et al 2016), more detailed results and discussions are presented in this article.…”

“…This article is an extension of our previous study (Shen, Chu, Cromieres, and Kurohashi 2016). In (Shen et al 2016), we only experimented on the Japanese-Chinese language pair in the scientific domain, and thus the language and task independency of our proposed method is not clear.…”

“…In (Shen et al 2016), we only experimented on the Japanese-Chinese language pair in the scientific domain, and thus the language and task independency of our proposed method is not clear. In this article, we thus conduct additional experiments on the English-Chinese language pair in the patent and Olympic domains.…”