A new alignment algorithm for parallel corpora of Japanese and Chinese
Yuhua Quan, Ying-hao Jin, Jingji Quan · 2011
To improve the alignment efficiency of sentence for parallel corpora, a new method is presented. It builds the non-grammar relation between source and target sentences by boolean transition, computes average match scores well and truly by boolean transition of sentences, and classifies all parallel sentences into 1:1, 1:n, n:1 and n:m by alignment relations. This method can not only increase the alignment efficiency, but also improve the accuracy rate of sentence alignment. Experiments on computer show that this new method is more adaptable and practicable.