Intelligent Tutoring Systems

Abstract

Intelligent Tutoring System (ITSs) implement many affordances of digital technologies that were articulated in the National Academy of Sciences, Engineering, and Medicine report How People Learn II. ITSs systematically respond to actions of the student (interactivity), present information contingent on the behavior, knowledge, and characteristics of the student (adaptivity), immediately provide feedback on the quality of the student's responses and how they could be improved (feedback), and generate hints, explanations, and other informational guides. Socratic tutoring, modeling-scaffolding-fading, reciprocal teaching, frontier learning, building on prerequisites, or diagnosis/remediation of deep misconceptions are some of the complex tutoring strategies that are highly regarded in the education and ITS communities. Conversational tutors converse with students in natural language and are very different from cognitive tutors and CBM-based tutors. ITS development has also largely followed a cottage industry approach where individual researchers build new prototype systems, essentially starting from scratch each time.

Topics & Concepts