New to Crowdmark? Check out our Getting Started Guide for instructors.
Sign In
Person on laptop holding bright lightbulb

Intelligent Tutoring Systems: Connecting AI and Education

Discussions of artificial intelligence (AI) in education (AIED) are often dominated by a futurist approach. Discourse tends to focus on the endless possibilities that the technology will one day offer, instead of highlighting the performance of contemporary applications. While the current use of AI is assuredly just scratching the surface of its potential, there exists a myriad of significant ways in which AI is currently being used in teaching and learning. Amongst these, perhaps one of the most prominent applications can be found with intelligent tutoring systems (ITS).

A recent article from researchers Shihui Feng and Nancy Law from the University of Hong Kong found that in the academic literature on AIED, ITSs represent the most frequently appearing research cluster over the past 10 years. Beyond an increasing research presence, ITSs have become entryways for many students into the world of AI. Understanding ITSs, their various manifestations, and seeing a snapshot of their present visibility in education will help to show the state of AI in education and help educators begin thinking about incorporating technology in teaching and learning.

What are Intelligent Tutoring Systems?

ITSs all conform to the same basics: online based teaching that can generate feedback with no human involvement. A form of adaptive learning technology, the core idea behind ITSs is constant—to support learning— but, the design of these human-less, tailored instructional programs has been the focus of ongoing debates since the 1960s. As explained by Nkambou, Bourdeau and Mizoguchi, the architecture of ITSs consists of four components.

  • The domain model functions as a knowledge database for how the system develops, administers, and evaluates questions.
  • The student model is responsible for collecting data about the student and making determinations about their state of knowledge.
  • The tutoring model ties the first two components together using the information collected to inform the learning process.
  • The interface component hosts the communication, which acts like a bridge between the components and the student. While some ITSs may have additional complexities, this is the essence of how they operate.

Intelligent Tutoring Systems in Teaching and Learning

While the design remains a point of contention, many ITSs are in operation today. For example, AutoTutor, developed out of the University of Memphis, functions through a naturalistic dialogue with the student. The system is capable of generating complex problems and assists the student towards finding the solution through a back-and-forth type of communicative interface. The platform is typically used for math and science, but from the earliest stages, computer literacy was the primary focus.

The founders have published widely on the effectiveness of AutoTutor as a tool for learning and memorization. The key finding to come out of research on AutoTutor’s efficacy is the significance of language—by extension, the interface component—as the program thrives by creating a constructive discourse between the AI tutor and student. This technology supports teaching and learning by allowing students to practice and internalize foundational concepts once they have learned them in class.

Another interesting profile in this discussion can be found with Ametros Learning, a Canadian company that caters towards business education with AI-simulated negotiations. The system offers a scenario in which students then engage with various characters via email that facilitates a realistic problem-solving experience.

While more traditional ITSs are hinged on instruction and feedback—essentially replacing a human tutor with AI—the Ametros Learning platform differs as it creates an experiential learning opportunity not found in classrooms. More than simply human replacement, this type of AI application targets an entirely new educational approach. This represents a shift towards more creative applications of ITSs and is a manifestation of AI’s potential.

The Future of Intelligent Tutoring Systems

ITSs are a way to see AIED in practice as a form of innovative classroom technology with potential for academic assistance. ITSs can offer a practical solution to the various obstacles students may have accessing a human tutor (cost, location, time, etc.). In some ways, ITSs could be used to complement academic support programs premised on creating accommodations for students. Moreover, ITSs have proven effective in assisting learners with clear solution-driven subjects as well as providing students with genuinely unique learning experiences that transcend what is typically offered with traditional methods. ITSs will not replace teachers, but rather, exist as a digital tool for the classroom that complements them in a myriad of ways to enrich educational practices and excite learners.

Interested in more on technology in education? Enjoy these other posts: