A novel Chinese-English on translation method using mix-language web pages

In this paper, we propose a novel Chinese-English organization name translation method with the assistance of mix-language web resources. Firstly, all the implicit out-of-vocabulary terms in the input Chinese organization name are recognized by a CRFs model. Then the input Chinese organization name is translated without considering these recognized out-of-vocabulary terms. Secondly, we construct some efficient queries to find the mix-language web pages that contain both the original input organization name and its correct translation. At last, a similarity matching and limited expansion based translation identification approach is proposed to identify the correct translation from the returned web pages. Experimental results show that our method is effective for Chinese organization name translation and can improve performance of Chinese organization name translation significantly.

1. Introduction

These Named entity (NE) translation is to translate a NE from one language to the other, which is very important to many Natural Language Processing (NLP) tasks, such as cross-language information retrieval and data-driven machine translation. According to NE types, NE translation usually can be classified into three main sub-tasks: person name (PER) translation, location name (LOC) translation, and organization name (ON) translation. And ON translation is attracting more and more research attention. Many previous works have been done on ON translation. Recently, researchers (Y.AI-Onaizan et al. 2002, Huang et al. 2005, Jiang et al. 2007, Yang et al. 2008, Yang et al. 2009, Ren et al. 2009, and so on) have focused on translating ON with the assistance of web resources. It is believed that there must be some mix-language web pages that contain both the input ON and its correct translation since there are large amount of web pages on the web. Thus ON translation problem is converted to following two problems. One problem is to find the mix-language web pages that contain both the input ON and its correct translation. The other problem is to identify the correct translation from the obtained mix-language web pages.