依赖关系(UML)
透视图(图形)
计算机科学
距离测量
公制(单位)
自然语言处理
远程教育
语言能力
人工智能
数学教育
心理学
工程类
运营管理
作者
Jinghui Ouyang,Jingyang Jiang,Haitao Liu
标识
DOI:10.1016/j.asw.2021.100603
摘要
Syntactic complexity is one of the key research foci in writing assessment. This study combines traditional syntactic complexity (SC) measures with newly-proposed dependency distance measures to better assess second language (L2) SC development. Based on a syntactically-annotated corpus of 400 compositions, we aim to investigate the extent to which the traditional SC measures and dependency distance measures can differentiate the writing proficiency of beginner, intermediate, and advanced learners. In terms of traditional SC measures, length-based measures fare better when pinpointing the different writing proficiency levels, but none of these can significantly differentiate all adjacent levels. As for dependency distance measures, the overall mean dependency distance can significantly discriminate all pairs of adjacent proficiency levels, serving as the best metric explored in our study. Moreover, dependency distance measures can further explain the findings of the traditional SC measures from the perspective of language processing.
科研通智能强力驱动
Strongly Powered by AbleSci AI