机器学习算法在翻译风格研究中的应用

发布时间：2018-03-10 17:02

本文选题：机器学习算法　切入点：信息增益法　出处：《外语教学》2017年05期 　论文类型：期刊论文

【摘要】：支持向量机是文本分析领域应用最多的机器学习算法之一,但其在翻译风格研究中的应用却并不多见。本研究基于语言计量特征,运用机器学习的方法对比王科一、张经浩两译者《傲慢与偏见》译本的翻译风格。首先,用信息增益算法获取两译者具有差异性的语言特征,然后基于这些特征构建文本分类器即支持向量机(SVM),对两译者的翻译文本进行自动分类,最后用十折交叉法对分类器的有效性进行验证。研究证明,机器学习算法分类准确率高,且具有很强的预测功能,有助于判定或预测译本或者译者的风格。与传统方法相比,该方法省时省力,过程可重复,结果精确可靠,便于处理大量陌生语料和海量特征,能发现文本间更细微的特征差异,为大数据时代的探索性翻译研究提供了便利,是翻译研究科学化的一种新的尝试。另外,本研究结合统计数据对差异性语言特征进行分析,分析结果与前期研究和研究者对文本的感受高度吻合,具有很强的可解释性。
[Abstract]:Support vector machine (SVM) is one of the most widely used machine learning algorithms in the field of text analysis, but its application in translation style research is rare. The translation style of Zhang Jinghao's translation of Pride and Prejudice. Firstly, the information gain algorithm is used to obtain the different linguistic features of the two translators. Then, based on these features, a text classifier, support vector machine (SVM), is constructed to classify the translated text automatically. Finally, the validity of the classifier is verified by using the 10% crossover method. It is proved that the machine learning algorithm has high classification accuracy. Compared with the traditional method, the method can save time and effort, the process can be repeated, the result is accurate and reliable, and it is convenient to deal with a large number of unfamiliar data and massive features. It is a new attempt to explore translation studies in big data's time, and it is a new attempt to make translation studies more scientific. In addition, this study combines statistical data to analyze the characteristics of different languages. The results are highly consistent with previous studies and researchers' perception of the text.
【作者单位】：西安交通大学外国语学院;
【基金】：教育部社科规划项目“中医汉英平行语料库的构建与应用研究”(项目编号:15YJC740127) 国家社科基金规划项目“译出与译入文本语言特征的计量风格学研究”(项目编号:17BYY007)的部分研究成果
【分类号】：H059

【相似文献】