Implementation of a Question Generation System to promote Critical Readings in STEM Education
keywords:
Question generation, STEM education, advanced NLP methods, information extraction, knowledge-acquisition
Description:
Setting the foundations for the generation of pedagogically relevant questions and for the representation of the domain specific knowledge. Building a Question Generation (QG) system that generates deep question relying on the most advanced NLP methods. Perform a series of evaluations to measure the appropriateness of the approach.
Objectives:
The main goal of the project is to generate content-dependent deep questions to support critical reading in the Biology Domain. First, a domain specific knowledge-base will be generated by automatically extending existing resources. After that, a QG system will be developed that, given a text and relying on distant supervision, will identify the relevant pieces (e.g. formalizing it as a graph matching problem) and link them to the knowledge-base. Some templates will be manually described to make the QG system able to generate questions about the information from the text.
Task:
WP 1: Domain modeling
Task 1.1: Knowledge acquisition
Task 1.2: Dataset generation
WP 2: Text to question process
Task 2.1: Information extraction (Text to concept)
Task 2.2: Question generation (Concept to question)
WP 3: Evaluation
References:
Jouault C. & Seta K. (2013) Wikipedia-Based Concept-Map Building and Question Generation, JSTAGE, Vol. 12, No.1, pp. 50-55, (2013)
Ming, L., Calvo, R., &Rus, V. (2014) Automatic Generation and Ranking of Questions for Critical Review, Educational Technology Society, 14 (3), pp. 16-27
Mostow J., & Chen W. (2009) Generating Instruction Automatically for the Reading Strategy of Self-Questioning. AIED2009, pp. 465-472
Olney, A.M., Graesser, A.C., & Person N.K. (2012) Question generation from concept maps. Dialogue &Discourse, 3(2), 75 – 99
Ming, L., Calvo, R., &Rus, V. (2014) Automatic Generation and Ranking of Questions for Critical Review, Educational Technology Society, 14 (3), pp. 16-27
Mostow J., & Chen W. (2009) Generating Instruction Automatically for the Reading Strategy of Self-Questioning. AIED2009, pp. 465-472
Olney, A.M., Graesser, A.C., & Person N.K. (2012) Question generation from concept maps. Dialogue &Discourse, 3(2), 75 – 99
Team:
Itziar Aldabe, Oier Lopez de Lacalle, Montse Maritxalar
Profile:
Computer scientist
File:
contact:
itziar.aldabe[at|abildua]ehu.eus
other:
Date:
2017