Not Logged In
You need to be logged in, be a member, and get a ticket to attend this session.
A systematic review of 50 years of automated question generation research
Questions play a fundamental role in education. Although constructing questions has been, and is generally conceived of as, a manual process, Automated Question Generation (AQG) is a burgeoning area of Intelligent CALL (Underwood, 1989) formed from a synergy of research strengths from Artificial Intelligence and Natural Language Processing. Over the past 25 years of CALL research, AQG has been used to supplement educational activities (e.g., answering practice questions) in large-scale courses (e.g., MOOCs) and advance computer assisted language testing (e.g., adaptive testing).
This presentation will first offer an overview of the AQG research across language learning and assessment studies that covers six key dimensions: 1) knowledge source for question generation (e.g., ontology, text), 2) generation method (e.g., template, syntax), 3) question type (e.g., WH-factual, cloze), 4) response format (e.g., open-ended, multichoice), reliability check (e.g., correlation with external exam scores) and 6) evaluation (e.g., expert, student). The primary aims of the presentation are to provide an overview of the AQG research as related to language learning and to discuss with the audience the benefits and drawbacks of using automated generated questions for classroom activities and assessments.