Auxiliary Writing System In The Filed Of Computational Linguistics

As the global integration process of constant development, academic research cann't act blindly, need to communicate with foreign academia. But as non-English is the mother tongue writers who write English academic articles, often encounter the situation that how to express an English sentence to make foreign counterparts to really understand their own research result, ultimately how to write real English academic field sentences.The quality of current machine translation is far from practical application and it is impossible to help the users to write the real English sentences.English-Chinese bilingual sentences searching system allows the users to input Chinese to retrieve example sentences what they want. While the sentence retrieved isn't necessarily identical with the sentence what the users want, the users can draw lessons from and study these sentences and combine their own experiences to express what they want in English. This paper try to integrate versatile Natural language processing technology that include sentence searching, machine translation, words similarity and collocations, research and implementate a English auxiliary writing system that specific to the users who write English academic articles.In the second chapter the paper investigates and surveys correlational research of two main technology of English auxiliary writing machine translation and sentence searching, analysis the advantages and disadvantages of various methods, choose the method that include statistical phrase-based translation and sentence searching based on distance. In the third chapter the paper research the generation method of term suggestion of user query based on English auxiliary writing system, including ordered synonyms recommended and common collocation recommended. we investigate and survey firstly correlational research of words similarity, present a words similarity calculation method in a certain context and combine with semantic dictionary WordNet to implement ordered synonyms recommended. Then we use monolingual word alignment method to extract collocation, and research the statistics and semantic combined method to calculate words similarity, implement the generation of the common collocation recommended in one-word query and two-word query. In the fourth chapter the paper introduce the system architecture, computational linguistics and implementation methods of sytem modules in English auxiliary writing system.