基于Word2vec的二语教学“基本形式库”构建方法初探  

Construction of a“Basic Form Corpus”for TCSOL Based on Word2vec

作　　者：杨苛鑫 庄会彬 杨牧 YANG Kexin;ZHUANG Huibin;YANG Mu(School of Culture and Communication,Shandong University;School of International Studies,Zhejiang University)

机构地区：[1]山东大学文化传播学院 [2]浙江大学外国语学院

出　　处：《国际汉语教学研究》2024年第3期76-84,共9页Journal of International Chinese Teaching

摘　　要：二语教学中,重视表达取向的“基本形式”观致力于构建一个“基本形式库”。本文以微博语料库为例,将其中高频词设置为检索词,依据Word2vec训练的词向量进行检索,围绕检索词查找近似词来构成(准)等义组,继而进一步确定该组的“基本形式”。本文初步提出了一种兼具可操作性和效率性的建设方法,作为人工建设“基本形式库”的辅助工具,并检索出了一部分基本形式(准)等义组作为前人研究的补充,为“基本形式”观理论进一步发展提供工具与思考。In second language teaching,the“Basic Forms”concept,which emphasizes semantic expression,aims to construct a“Basic Form Corpus”.Taking the Weibo corpus as an example,this paper sets high-frequency words as search items and employs Word2vec-trained word vectors for retrieval.The(quasi-)synonym group is formed by searching for synonyms related to search items,further establishing the“Basic Forms”of the group.The paper proposes a preliminary method that combines operability and efficiency.As an auxiliary tool for constructing the“Basic Form Corpus”,a subset of the basic form(quasi-)synonym group is retrieved to supplement previous studies,offering some suggestions for further research of the“Basic Form”approach.

关 键 词：基本形式 Word2vec 词向量 (准)等义组 

分 类 号：H195[语言文字—汉语]