摘要
英语长句句式结构复杂,对机器翻译过程中的语义关联性分析产生干扰,影响机器翻译质量。为此,设计一种基于概念层次网络(HNC)理论的英语长句语义切分全局翻译优化方法。预处理英语长句特征语义块,构建句子逻辑与序列;基于HNC理论的句子语义切分算法切分英语长句,获取短句信息熵数据;采用自主适应匹配语义的非线性谱特征翻译短句,识别待翻译文本语义信息特征,调整翻译后的短句顺序,对其进行合成获得全局翻译结果。实验结果表明,该方法的切分准确率较高,可精准识别英语长句,具有较高的机器翻译质量。
The complex structure of long English sentences interferes with the semantic correlation analysis in the process of machine translation,affecting the quality of machine translation.Therefore,an optimiza⁃tion method of semantic segmentation of long English sentences based on Hierarchical Network of Concepts(HNC)theory is designed.Feature semantic blocks of long English sentences are preprocessed to construct sentence logic and sequence.Sentence semantic segmentation algorithm based on HNC theory is used to segment long English sentences and obtain information entropy data of short sentences.Short sentences are translated with nonlinear spectral features that can automatically adapt to matching semantics,identify the features of semantic information of the text to be translated,adjust the order of translated short sentences,and then synthesize them to obtain global translation results.The experiment results show that the proposed method has a high segmentation accuracy,can accurately identify long English sentences,and has a high quality of machine translation.
作者
牛小青
NIU Xiao-qing(Department of Foreign Languages,Xi’an Jiaotong University,City College,Xi’an 710018,China)
出处
《信息技术》
2024年第8期121-126,共6页
Information Technology
基金
陕西省教育科学“十三五”规划2017年度课题(SGH17H393)。
关键词
语义切分
英语长句
机器翻译
HNC理论
句子切分算法
semantic segmentation
long English sentences
MT
HNC theory
sentence segmentation al⁃gorithm
