Edited by Sandra Götz and Joybrato Mukherjee
[Studies in Corpus Linguistics 92] 2019
► pp. 29–48
To automated generation of test questions on the basis of error annotations in EFL essays
A time-saving tool?
The paper presents computer tools that were designed to increase the pedagogical value of a learner corpus for teachers and learners of English as a foreign language. The corpus used in the study is a collection of essays written by Russian learners of English – students from four different cities in Russia, studying at different Bachelor programmes at the Higher School of Economics. The paper describes how the tools we offer make use of manually annotated errors for the process of automated generation of questions for English tests, what editing the generated tests requires, and in the final section there is a discussion to what degree this process appears to be useful for EFL instructors.
- 1.1Previous research in corpus annotation
- 1.2Description of the learner corpus and annotation practices adopted in it
- 2.Benefits of error annotations in REALEC for teaching and learning English
- 2.1Pedagogical directions in using a learner corpus in university EFL classes
- 2.2REALEC English Test Maker: Related works, the main principles and stages of work of the program
- 2.3REALEC English Test Maker: Special cases
- 2.4RETM (REALEC English Test Maker): Editing automatically generated questions