Modeling Student Response Times: Towards Efficient One-on-one Tutoring Dialogues

LUCIANA BENOTTI; JAYADEV BHASKARAN; DAVID LANG; SIGTRYGGUR KJARTANSSON

Proceedings of the Conference on Empirical Methods for Natural Language Processing

Association for Computational Linguistics

Año: 2018; p. 121 - 131

In this paper we investigate the task ofmodeling how long it would take a studentto respond to a tutor question during a tutoringdialogue. Solving such a task hasapplications in educational settings suchas intelligent tutoring systems, as wellas in platforms that help busy human tutorsto keep students engaged. Knowinghow long it would normally take a studentto respond to different types of questionscould help tutors optimize their owntime while answering multiple dialoguesconcurrently, as well as deciding when toprompt a student again. We study thisproblem using data from a service that offerstutor support for math, chemistry andphysics through an instant messaging platform.We create a dataset of 240K questions.We explore several strong baselinesfor this task and compare them with human performance.