The traditional approach to presenting idioms relies mainly on teachers’ or materials writers’ judgement, one-by-one and quite incidentally; and the existing teaching materials and references for idioms are mostly intuition-based. However, a more recent approach to better teaching and learning idioms is to present them under categories of their common themes and topics. Corpus linguistics can be of much contribution through helping the design and development of more authentic and systematic materials using comprehensive corpora which are typically the best representatives of the target language. In this connection, the present study aimed at searching for the thematic index of 1506 idioms under 81 categories at the end of the Oxford Dictionary of Idioms in the largest freely available corpus, i.e. the Corpus of Contemporary American English (COCA), composed of more than 520 million words. To this end, we used a manuscript in PHP written by a professional computer programmer especially for this purpose. The findings yielded a list of idioms sorted based on their frequencies under their theme-based categories. To focus on the more frequently-used idioms of various themes in real contexts, materials designers, teachers, and learners of English can benefit from this idiom list in textbooks and classroom activities.

中文翻译：

---

基于主题语料库的当代美国英语语料库中的成语研究

呈现习语的传统方法主要依靠老师或材料作家的判断，而且是一种偶然的方式。现有的成语教材和参考文献大多是基于直觉的。但是，最近一种更好的教学习语的方法是将它们按共同主题和主题的类别进行介绍。语料库语言学可以通过使用全面的语料库（通常是目标语言的最佳代表）帮助设计和开发更真实和系统的材料而做出巨大贡献。在这方面，本研究旨在在最大的免费可用语料库即《当代美国英语语料库》（COCA）中，在《牛津成语词典》末尾的81个类别中搜索1506个成语的主题索引，由5.2亿多个单词组成。为此，我们特别使用了由专业计算机程序员编写的PHP手稿。调查结果列出了成语列表，这些成语根据其在基于主题的类别下的出现频率进行了排序。为了专注于实际情况下各种主题的更常用成语，材料设计人员，教师和英语学习者可以从教科书和课堂活动中的该成语列表中受益。