Identify Bilingual Patterns and Phrases from a Bilingual Sentence Pair

Abstract This paper presents a method for automatically identifying bilingual grammar patterns and extracting bilingual phrase instances from a given English- Chinese sentence pair. In our approach, the English-Chinese sentence pair is parsed to identify English grammar patterns and Chinese counterparts. The method involves generating translations of each English grammar pattern and calculating translation probability of words from a word-aligned parallel corpora. The results allow us to extract the most probable English-Chinese phrase pairs in the sentence pair. We present a prototype system that applies the method to extract grammar patterns and phrases in parallel sentences. An evaluation on randomly selected examples from a dictionary shows that our approach has reasonably good performance. We use human judge to assess the bilingual phrases generated by our approach. The results have potential to assist language learning and machine translation research. Keywords: Pattern Grammar, Phrase Translation, Word Alignment 1 Introduction Verb phrases are prominent components of any sentence. If we can correctly extract English- Chinese bilingual grammar patterns and phrases in a bilingual sentence pair, it will be helpful for English and Chinese language learners and machine translation systems. These results can be used to demonstrate the synchronous structure of parallel sentences. However, Chinese sentence parsing technology is still immature, which leads to difficulties in obtaining grammar patterns and phrases in Chinese sentences. There are still many problems with existing Chinese parsers. To extract Chinese patterns and phrases more accurately, we utilize an English sentence parser which is much more mature than Chinese parsing technology to parse English sentences and extract English patterns and phrases. We also utilize some statistical methods to estimate translation probability of word pairs using bilingual corpora. Then, we extract Chinese patterns and phrases more accurately by finding counterparts of English patterns and phrases with statistical results from bilingual corpora. We present a system that returns bilingual verb patterns and phrases of a bilingual sentence pair. Our system identifies phrases in sentences by using the pattern table and translation probability model created in advance. The pattern table created by parsing English sentences and calculating counterparts in bilingual parallel corpus with word alignment. The translation probability model is created by the alignment probability of each English and Chinese word pair with consideration of not only the word itself but also related words. At the runtime, our system starts with an English-Chinese bilingual sentence pair submitted by the user. The system parses the English sentence and extracts English patterns and phrases, and retrieves the counterpart of the pattern from the table created in advance to be Chinese pattern. Then, the system extracts the counterpart of words in the English phrase by the probability model created in advance. Finally, The system combines the Chinese pattern and the counterpart of words in the English phrase to form a Chinese phrase. The system can assist language learning or be used to generate training data for machine Identify Bilingual Patterns and Phrases from a Bilingual Sentence Pair Yi-Jyun Chen Department of Computer Science National Tsing …

Content truncated to 500 words