Automatic classification of teacher feedback and its potential applications for EFL writing

This paper presents and discusses the initial data from a project that aims to develop a system for automatic tracking of student responses to teacher feedback in draft revision. One main purpose of the project is to design and implement a method for automatic classification of teacher feedback on students’ draft essays in the EFL context. In this paper, we propose the automatic classification method and evaluate its performance in terms of accuracy. Our findings show that an accuracy of over 96% was achieved when classifying teacher feedback using the proposed method. They also show that the classification results could be analysed with other sets of data such as assessment grades to help teachers reflect on their use of feedback types and refine their feedback practice. This study can provide a basis for future research into automatic analysis of the impact of various feedback types on student revision. Copyright © 2017 Asia-Pacific Society for Computers in Education.