Bilingual dictionary based sentence alignment for Chinese English bitext

Bitext is a rather “hot” issue among current natural language processing and automatic sentence alignment is the first step towards bitext processing. Following the shift form previous knowledge-poor approach (pure length-based) to current some-what knowledge-rich approach, this paper suggests a bilingual dictionary based sentence alignment method for Chinese English bitext and realizes it through dynamic programming. Experiment on HIT bitext shows that this method has achieved an accuracy as high as 95%, and therefore is worthy of further exploring.