A Statistical Model-Based Approach to NE Translation

Named Entities (NE) is widely used in various research areas such as machine translation, cross-language information retrieval, bilingual question answering, bilingual lexicon construction and so on. Although in recent years scientists have done great efforts to identify and translate NE, aligning named entities in bilingual documents is still a complex problem. In this paper, we propose a method to align bilingual named entities in parallel corpora. The results show that the proposed methodology has got good performance, and it can be applicable to a wide range of natural language processing.