The Construction of Chinese Multi-dimensional Learner Corpus:YACLC
WANG Yingying,KONG Cunliang,YANG Liner,HU Renfen,YANG Erhong,SUN Maosong
DOI: https://doi.org/10.16499/j.cnki.1003-5397.2023.01.005
IF: 3.6
2023-01-01
Applied Linguistics
Abstract:Guided by the theory and the methods of Contrastive Interlanguage Analysis and intelligent computer-assisted writing, this paper constructs a large-scale, high-quality, document-level,multi-dimensional annotated Chinese learner corpus, Yet Another Chinese Learner Corpus(YACLC).YACLC designs a multi-dimensional informative annotation guideline, including minimal edit, fluency edit, sentence acceptability, and context dependence. Then YACLC annotates 2,421 Chinese learner texts of language usage scenarios with 32,124 sentences using a crowdsourcing strategy, to obtain 331,292minimal edit annotations and 137,708 fluency edit annotations. The construction of YACLC not only solves the problems of closed data resources, single annotation and lacking of fluency dimension of the Chinese learner corpus, but also supports and extends the comparative analysis between the learner language and the two reference language variants to reveal the laws of second language acquisition.