基于语义网络的英语机器翻译模型设计与改进
|
摘要
针对传统基于规则的机器翻译模型存在英语翻译结果不够精确、难以准确描述词语间关系的弊端,设计并改进基于语义网络的英语机器翻译模型。该模型采用基于向量混合的短语合成语义统计英语机器翻译方法,在翻译相似度模型中,采用余弦相似度的方法获取两个向量的语义相似度,经过带权向量加法的计算极易辨别两个相似向量的不同之处,获取精准的英语翻译结果,对句子实施权值训练获取构成句子的主要短语,保证翻译结果归纳出句子的中心思想。改进基于语义网络的英语机器翻译模型,针对用户需求引入大数据的同时让语言学家参与到机器翻译的过程中,使得英语翻译结果既能独立进行语义表达,又能准确描述词语间关系。实验结果表明,所设计的模型能够精准高效地进行英语翻译。
In allusion to the deficiencies existing in the traditional rule-based machine translation model for its inaccurate English translation results and difficulty to accurately describe the relationship between words,an English machine translation model based on semantic network is designed and improved. In the model,the phrase semantic synthesis statistical English machine translation method based on vector hybrid is adopted. In the translation similarity degree model,the cosine similarity degree method is adopted to obtain the semantic similarity degree of two vectors. The differences between two similar vectors are very easy to be discriminated after addition calculation of weighted vectors,so as to obtain accurate English translation results.The weight training is conducted for sentences to obtain the main phrases that constitute sentences,so as to ensure that the central idea of the sentence is summarized in translation results. In the improved English machine translation model based on semantic network,big data is introduced to meet users′ needs and linguists are invited to participate in the machine translation process,so that not only can semantic expressions be independently conducted,but also the relationship between words can be accurately described in English translation results. The experimental results show that the designed model can conduct an accurate and efficient English translation.
In allusion to the deficiencies existing in the traditional rule-based machine translation model for its inaccurate English translation results and difficulty to accurately describe the relationship between words,an English machine translation model based on semantic network is designed and improved. In the model,the phrase semantic synthesis statistical English machine translation method based on vector hybrid is adopted. In the translation similarity degree model,the cosine similarity degree method is adopted to obtain the semantic similarity degree of two vectors. The differences between two similar vectors are very easy to be discriminated after addition calculation of weighted vectors,so as to obtain accurate English translation results.The weight training is conducted for sentences to obtain the main phrases that constitute sentences,so as to ensure that the central idea of the sentence is summarized in translation results. In the improved English machine translation model based on semantic network,big data is introduced to meet users′ needs and linguists are invited to participate in the machine translation process,so that not only can semantic expressions be independently conducted,but also the relationship between words can be accurately described in English translation results. The experimental results show that the designed model can conduct an accurate and efficient English translation.
引文
[1]刘宇鹏,马春光,张亚楠.深度递归的层次化机器翻译模型[J].计算机学报,2017,40(4):861-871.LIU Yupeng,MA Chunguang,ZHANG Yanan.Hierarchical machine translation model based on deep recursive neural network[J].Chinese journal of computers,2017,40(4):861-871.
[2]李响,南江,杨雅婷,等.泛化语言模型在汉维机器翻译中的应用[J].计算机应用研究,2014,31(10):2994-2997.LI Xiang,NAN Jiang,YANG Yating,et al.Application of generalization language model in Chinese-Uyghur machine translation[J].Application research of computers,2014,31(10):2994-2997.
[3]ZHANG J,LIU S,LI M,et al.Towards machine translation in semantic vector space[J].ACM transactions on Asian and lowresource language information processing,2015,14(2):9.
[4]MUZAFFAR S,BEHERA P,NATH G.A Pāniniān framework for analyzing case marker errors in English-Urdu machine translation[J].Procedia computer science,2016,96(C):502-510.
[5]惠浩添,李云建,钱龙华,等.一个面向信息抽取的中英文平行语料库[J].计算机工程与科学,2015,37(12):2331-2338.HUI Haotian,LI Yunjian,QIAN Longhua,et al.A ChineseEnglish parallel corpus for information extraction[J].Computer engineering and science,2015,37(12):2331-2338.
[6]薛征山,张大鲲,王丽娜,等.改进机器翻译中的句子切分模型[J].中文信息学报,2017,31(4):50-56.XUE Zhengshan,ZHANG Dakun,WANG Lina,et al.An improved sentence segmentation model for machine translation[J].Journal of Chinese information processing,2017,31(4):50-56.
[7]ROMOOZI M,FATHY M,BABAEI H.A content sharing and discovery framework based on semantic and geographic partitioning for vehicular networks[J].Wireless personal communications,2015,85(3):1583-1616.
[8]王俊华,左祥麟,左万利.基于证据理论的单词语义相似度度量[J].自动化学报,2015,41(6):1173-1186.WANG Junhua,ZUO Xianglin,ZUO Wanli.Word semantic similarity measurement based on evidence theory[J].Acta automatica sinica,2015,41(6):1173-1186.
[9]MALLAT S,MOHAMED M A B,HKIRI E,et al.Semantic and contextual knowledge representation for lexical disambiguation:case of Arabic-French query translation[J].Journal of computing&information technology,2014,22(3):191-215.
[10]BOULARES M,JEMNI M.Learning sign language machine translation based on elastic net regularization and latent semantic analysis[J].Artificial intelligence review,2016,46(2):145-166.
[2]李响,南江,杨雅婷,等.泛化语言模型在汉维机器翻译中的应用[J].计算机应用研究,2014,31(10):2994-2997.LI Xiang,NAN Jiang,YANG Yating,et al.Application of generalization language model in Chinese-Uyghur machine translation[J].Application research of computers,2014,31(10):2994-2997.
[3]ZHANG J,LIU S,LI M,et al.Towards machine translation in semantic vector space[J].ACM transactions on Asian and lowresource language information processing,2015,14(2):9.
[4]MUZAFFAR S,BEHERA P,NATH G.A Pāniniān framework for analyzing case marker errors in English-Urdu machine translation[J].Procedia computer science,2016,96(C):502-510.
[5]惠浩添,李云建,钱龙华,等.一个面向信息抽取的中英文平行语料库[J].计算机工程与科学,2015,37(12):2331-2338.HUI Haotian,LI Yunjian,QIAN Longhua,et al.A ChineseEnglish parallel corpus for information extraction[J].Computer engineering and science,2015,37(12):2331-2338.
[6]薛征山,张大鲲,王丽娜,等.改进机器翻译中的句子切分模型[J].中文信息学报,2017,31(4):50-56.XUE Zhengshan,ZHANG Dakun,WANG Lina,et al.An improved sentence segmentation model for machine translation[J].Journal of Chinese information processing,2017,31(4):50-56.
[7]ROMOOZI M,FATHY M,BABAEI H.A content sharing and discovery framework based on semantic and geographic partitioning for vehicular networks[J].Wireless personal communications,2015,85(3):1583-1616.
[8]王俊华,左祥麟,左万利.基于证据理论的单词语义相似度度量[J].自动化学报,2015,41(6):1173-1186.WANG Junhua,ZUO Xianglin,ZUO Wanli.Word semantic similarity measurement based on evidence theory[J].Acta automatica sinica,2015,41(6):1173-1186.
[9]MALLAT S,MOHAMED M A B,HKIRI E,et al.Semantic and contextual knowledge representation for lexical disambiguation:case of Arabic-French query translation[J].Journal of computing&information technology,2014,22(3):191-215.
[10]BOULARES M,JEMNI M.Learning sign language machine translation based on elastic net regularization and latent semantic analysis[J].Artificial intelligence review,2016,46(2):145-166.