Mining Large-scale Comparable Corpora from Chinese-English News Collections

In this paper, we explore a CLIR-based approach to construct large-scale Chinese-English comparable corpora, which is valuable for translation knowledge mining. The initial source and target document sets are crawled from news website and standardized uniformly. Keywords are extracted from the source document firstly, and then the extracted keywords are translated and combined as query words through certain criteria to retrieve against the index created using target document set. Meanwhile, the mapping correlations between source and target documents are developed according to the value of similarity calculated by the retrieval tool. Two methods are evaluated to filter the comparable document pairs so as to ensure the quality of the comparable corpora. Experimental results indicate that our approach is effective on the construction of ChineseEnglish comparable corpora.

• Type: Journal
• Year: 2010
• Where: COLING
• Authors: Degen Huang, Lian Zhao, Lishuang Li, Haitao Yu
• Keywords: Chinese-English Comparable Corpora, COLING 2010, Comparable Corpora, Computational Linguistics, Target Documents