Speech and Language Technology in Education (SLaTE 2013)

Grenoble, France
August 30-September 1, 2013

Filtering-based Automatic Cloze Test Generation

Kyusong Lee (1), Soo-Ok Kweon (2), Hae-Ri Kim (3), Gary Geunbae Lee (1)

(1) Department of Computer Science and Engineering; (2) Division of Humanities and Social Sciences;
Pohang University of Science and Technology, Korea
(3) Department of English Education, Seoul National University of Education, Korea

We propose a method to generate high-quality cloze test questions using a computational approach. Previous methods for automatic cloze test generation have contained some problems; specifically, there can be multiple correct answers. We found that approximately 50% of the generated answers have such errors with previous methods, which requires human post-editing was necessary in previous research. We propose an N-gram filtering method that can detect the answer to a given question. We compare the errors of the generated questions before and after applying the filtering methods. We found that our filtering method can select quality distractors by reducing errors in the generated questions. Moreover, when we generate cloze tests using semantic similarity, non-native speakers are very hard to answer the questions.

Index Terms: Cloze Test Generation, Sentence Completion Task, Vocabulary Question Generation

Full Paper

Bibliographic reference.  Lee, Kyusong / Kweon, Soo-Ok / Kim, Hae-Ri / Lee, Gary Geunbae (2013): "Filtering-based automatic cloze test generation", In SLaTE-2013, 72-76.