From the earliest examples of Pressey’s teaching machines (1924) to the immersive, game-based virtual worlds of today, the idea to simulate the activities of expert teachers and tutors has led to profound advances in educational technology. Educational applications are consistently identified by artificial intelligence (AI) researchers as an important goal. For example, inspired by Neil Stephenson’s science fiction novel The Diamond Age (1995), AI pioneer Marvin Minsky and his colleagues recently highlighted key ideas involved in building an intelligent tutoring system (ITS):
…we could try to build a personalized teaching machine that would adapt itself to someone’s particular circumstances, difﬁculties, and needs. The system would carry out a conversation with you, to help you understand a problem or achieve some goal. You could discuss with it such subjects as how to choose a house or car, how to learn to play a game or get better at some subject, how to decide whether to go to the doctor, and so forth. It would help you by telling you what to read, stepping you through solutions, and teaching you about the subject in other ways it found to be effective for you. Textbooks then could be replaced by systems that know how to explain ideas to you in particular, because they would know your background, your skills, and how you best learn. (Minsky et al., 2004, p. 122)
Decades of evidence suggest that systems incorporating rich representations of task domain and pedagogical knowledge are able to produce larger gains in learning than less sophisticated counterparts and classroom instruction. These benefits are usually attributed to the ability of AI-based educational systems to:
- track the “mental steps” of the learner and underlying goal structure of problem solving tasks (Anderson et al., 1995)
- diagnose misconceptions and estimate the learner’s understanding of the domain (VanLehn, 1988)
- provide timely guidance, feedback and explanations (Shute, 2008)
- promote productive learning behaviors, such as self-regulation, self-monitoring, and self-explanation (Azevedo & Hadwin, 2005)
- prescribe learning activities at the right level of difficulty and with the most appropriate content (VanLehn, 2006)
Essentially every AI technique – natural language processing, uncertain reasoning, planning, cognitive modeling, case-based reasoning, machine learning and more – has been applied to these specific challenges (Woolf, 2009).
AI research in education is not limited to simulating expert human tutors and teachers, however. For example, the use of physiological monitoring technologies (e.g., skin conductance, posture) is helping researchers understand the role of emotions in learning and develop new models of pedagogical intervention. Further, work on teachable agents leverages the idea of reciprocal teaching by having the (human) student take on the role of teacher. Here, AI techniques are used in a variety of ways, such as simulating human communication, learning, and emotions. A final example is the use of narrative learning environments that seek to provide customized experiential learning opportunities and better maintain learner engagement.
AI enables educational technologies to better understand, adapt to, and help learners in various ways. ITSs already consistently outperform untrained tutors and are now approaching the effectiveness of expert tutors. The potential to provide individualized tutoring support to every learner in every situation would revolutionize education. The idea that homework might be viewed not as required drudgery, but as an opportunity to succeed and demonstrate understanding is compelling. Further, the diagnostic capabilities of such systems can be used to provide teachers with detailed information about their students – what they understand and where they are struggling.
Intelligent tutoring represents a key success story for AI: approximately 500,000 middle- and high-school students use math tutors every year. Although researchers continue to work to improve the effectiveness of ITSs, AI-based educational technologies will continue to evolve alongside other major technological developments, such as social media, video games, and mobile technologies. These technologies enable powerful new ways to support learners across contexts and over extended periods of time. The vision of personalized lifelong learning support via ubiquitous computing and lifelong learner modeling suggests that lines between formal and informal learning will continue to blur (Kay, 2008). The need for deeper AI models for diagnosing and understanding learners (e.g., in social contexts), modeling their desires and interests, pointing them to resources, and supporting them in productive learning behaviors (e.g., self-monitoring and reflection) in these new contexts will be essential research goals.
Grand challenges as identified by key organizations and working groups:
- The Computing Research Association (CRA) identified the goal to Provide a Teacher for Every Learner as one of its five Grand Research Challenges.
- The United Kingdom Computing Research Committee (UKCRC) identified two grand challenges related to learning: Learning for Life (GC8) and Memories for Life (GC3).
- The National Academy of Engineering identified Advance Personalized Learning as a grand challenge.
Specific challenges for building and transitioning computer-based, intelligent learning environments:
- Cost to build, author content. Authoring systems.
- Transition and integration into educational practice (“stakeholder” involvement)
- Expand impacts beyond math and science instruction; ill-defined domains; informal learning contexts
- Expanding the communicative bandwidth of educational technologies
- User sensing and classification of user emotional states allows systems to track frustration, confusion, and engagement.
- Animated pedagogical agents increase social presence and enable new styles of feedback through nonverbal channels.
- Speech recognition and dialogue
Dissenting views / critiques
- The use of ubiquitous technologies and the pursuit of lifelong learner models will increase the salience of privacy and user control of their own data, even when used only for educational purposes (Kay, 2008).
- Search for lower cost methods for achieving similar learning outcomes.
- Are we planting the seeds to eventually replace teachers?
If we individualize educational approaches are learners going to have opportunities to learn and work together?
Top related areas in cyberlearning
Educational Data Mining
Games and Virtual Worlds
Full-body and Intelligent Interfaces
AAAI topics page on education (from the Association for the Advancement of Artificial Intelligence)
International Artificial Intelligence in Education Society
A Roadmap for Learning Technology (from the CCC, CRA, and NSF)]
Anderson, J.R., Corbett, A.T., Koedinger, K.R., & Pelletier, R. (1995). Cognitive tutors: Lessons learned>. Journal of the Learning Sciences 4(2): 167-207.
Azevedo, R. and A. Hadwin (2005). Scaffolding Self-regulated Learning and Metacognition – Implications for the Design of Computer-based Scaffolds. Instructional Science 33(5): 367-379.
Collins, A. & R. Halverson (2009). Rethinking education in the age of technology: the digital revolution and schooling in America. New York, Teachers College Press.
Kay, J. (2008). Lifelong learner modeling for lifelong personalized pervasive learning. IEEE Transactions on Learning Technologies 1(4): 215-228.
Minsky, M.L., Singh, P., & Sloman, A. (2004). The St. Thomas Common Sense Symposium: Designing Architectures for Human-Level Intelligence. AI Magazine 25(2): 113-125.
Pressey, S.L. (1927). A machine for automatic teaching of drill material. School and Society, 25(645): 549-552.
Shute, V.J. & Psotka, J. (1996). Intelligent tutoring systems: Past, present, and future. Handbook for research for educational communications and technology. D. H. Jonassen (Ed.). New York, NY, Macmillan: 570-599.
Shute, V. J. (2008). Focus on Formative Feedback. Review of Educational Research 78(1): 153-189.
Stephenson, N. (1995). The Diamond Age: Or, a Young Lady’s Illustrated Primer. Spectra.
VanLehn, K. (1988) Student modeling. In M. Poison & J. Richardson (Eds.), Foundations of Intelligent Tutoring Systems. Hillsdale, NJ: Erlbaum.
VanLehn, K. (2006). The Behavior of Tutoring Systems. International Journal of Artificial Intelligence in Education 16(3): 227-265.
Woolf, B.P. (2009). Building Intelligent Interactive Tutors: Student-centered Strategies for Revolutionizing E-learning. Amsterdam, Netherlands, Morgan Kaufmann.