Using computer-like rules to give automatic grammatical written corrective feedback a case study of structural particles“的”, “地” and “得”

Mandarin speakers often confuse three structural particles “的”, “地” and “得” in their formal writings and Chinese teachers often correct these errors in students' essays. To reduce Chinese teachers' laborious workload, the present research aims to investigate how to use technology to efficiently correct grammar errors with feedback. All three structural particle instances were analyzed in 50 primary one level HK students' essays. Their grammaticality judgments were based on three rules derived from literature: (1) A+的+B. B is always a Noun. A is used to modify B; (2) A+地+B. B is always a Verb. A is used to modify B to indicate how; (3) A+得+B. A is always a Verb. B is used to modify A to indicate how. 700 out of 804 instances of “的”, “地” and “得” were identified as structural particles. By using automatic Chinese parsing software NLPIR-ICTCLAS (see Zhang, 2018), 80% could be judged easily and accurately while the rest required a deeper analysis, which indicates that automatic Part of Speech (PoS) judgment is quite useful and helpful to Chinese teachers' grammatical feedback on these three structural particles in students' essays. Difficulties and problems in grammaticality judgment from a deep level were further analyzed. Corresponding solutions and directions for future research were also discussed. This research can shed light on how to use Part of Speech (PoS) parsing software to facilitate automatic grammatical written corrective feedback and how technology can enhance language learning and teaching in a Chinese language education context. Copyright © ICCE 2020 - 28th International Conference on Computers in Education, Proceedings. All rights reserved.