A System for Generating Cloze Test Items from Russian-Language Text
Andrey Kurtasov · Recent Advances in Natural Language Processing · 2013
This paper studies the problem of automated educational test generation. We describe a procedure for generating cloze test items from Russian-language text, which consists of three steps: sentence splitting, sentence filtering, and question generation. The sentence filtering issue is discussed as an application of automatic summarization techniques. We describe a simple experimental system which implements cloze question generation and takes into account grammatical features of the Russian language such as gender and number.