Research on the semantic knowledge base in Chinese-English machine translation

Knowledge bases in machine translation (MT) systems have proved successful in some constrained domains, but have not scaled up for two reasons. One is that the building of knowledge base (KB) is painstakingly handcrafted from scratch, and the other is that the most KBs for machine translation (MT) lack supports from powerful theory system based on semantic understanding. This paper focuses on the building of semantic knowledge base (SKB) guided by the Concept of Hierarchical Network (HNC) theory which is suitable for machine translation. Besides bilingual general attributes, the semantic attributes at all levels are described in a word such as concept category, semantic representation, and sentence category and concept relation. By doing this, we try to solve the semantic mapping problems between Chinese and English at the level of word, chunk and sentence. The SKB has been used both in the analysis of the source language and target language translation process. The accuracy of translation system based on the SKB has increased considerably.