基于改进聚类算法的英文翻译误差校正系统

English Translation Error Correction System Based on Improved Clustering Algorithm

为提升英文翻译正确率,精准识别短语语义,提出一种基于改进聚类算法的翻译误差校正系统。依靠夹角余弦计算语义相似度并加以分类,确定语义权重,利用四元集群对翻译前后结果似然运算,使用Logistics模型提取存在误差的语义特征,加入翻译概念集匹配语境,同时融入吸引子校正误差,通过搜索模块响应用户输入指令,在DOS系统C语言开发软件下获取存在翻译误差的英文语段,并将结果传输至校正计算内,以供英文翻译误差校正算法进行处理。实验结果表明,所提系统误差识别精准度高、校正准确,能妥善处理英文翻译存在的误差问题。

In order to improve the accuracy of English translation and accurately identify phrase semantics,a translation error correction sys-tem based on improved clustering algorithm is proposed.The semantic similarity is calculated and classified by cosine of includ-ed angle,and the semantic weight is determined.The four element cluster is used to calculate the likelihood of the results before and after translation.The semantic features with errors are extracted by using the logistics model,and the translation concept set is added to match the context.At the same time,the attractor is added to correct the errors.The search module responds to the us-er's input instructions,Under the C language development software of DOS system,the English segments with translation errors are obtained,and the results are transmitted to the correction calculation for processing by the English translation error correction algorithm.The experimental results show that the proposed system has high accuracy in recognition and correction,and can prop-erly deal with the errors in English translation.