论文部分内容阅读
能够有效反映语篇质量的语言特征主要包括语篇连贯性、词汇复杂性、句法复杂性以及语法准确性等,语篇连贯性是高质量作文的核心特征之一。本文基于Michael Hoey的词汇衔接理论,利用Word Net语义网络提取学生作文中与语篇连贯相关的文本特征;以人工连贯性评分为标准,确定与语篇连贯性有关的预测因子,尝试构建中国英语学习者书面语语篇连贯自动评价模型。模型验证结果表明,本研究构建的语篇连贯评价模型能够应用于中国英语学习者书面语语篇连贯性的自动评价,机器评分与人工评分在一致性方面已达到较高水平。
The language features that can effectively reflect the quality of discourse include discourse coherence, vocabulary complexity, syntactic complexity and grammatical accuracy. Discourse coherence is one of the core features of high quality composition. Based on Michael Hoey’s lexical cohesion theory, this paper uses WordNet semantic network to extract text features related to coherence of discourse in student essay. Using artificial coherence score as a criterion, we determine the predictive factors related to discourse coherence, and try to construct Chinese English Learner - Written Discourse Coherence Automatic Evaluation Model. The model verification results show that the coherence assessment model constructed in this study can be applied to the automatic assessment of Chinese EFL learners’ coherence of written language discourse, and the machine scores and manual scores have reached a high level of consistency.