回复: 求中韩平行语料库
Although many parallel corpora have been built, such as the Canadian Hansards (Gale and Church, 1991), the Europarl corpus (Koehn, 2005), the Arabic-English and English-Chinese parallel corpora used in the NIST Open MT Evaluation1, few parallel corpora exist for other language pairs, such as Chinese-Japanese, Japanese-Korean, Chinese- French or Japanese-German. Since parallel data are critical resources for building multilingual applications (e.g. machine translation and cross-lingual information retrieval), the shortage of parallel corpora would be a major limitation for multilingual applications to achieve higher performance.
