Artificial Intelligence in Online Education

Artificial Intelligence has been around for a long time. It was first coined in by John McCarthy in 1955 “McCarthy coined the term “AI” in 1955 in connection with a proposedsummer workshop at Dartmouth College, which many of the world’s leading thinkers in computing attended. (Computer History Museum, 2015). McCarthy defines Artificial Intelligence as “the science and engineering of making intelligent machines” in his article titled, “What is Artificial Intelligence?” (2007)

The “intelligence” in business intelligence and the “smart” in smart systems for the home, office and school are key indicators of the use of Artificial Intelligence. It can also be seen in many educational applications such as games, simulations, tutoring and collaborative systems. With the large amount of electronic data available today it easy to create a knowledge base. A knowledge base is one of the cornerstones of artificial intelligence in machine learning.

The Challenge

The challenges with Artificial Intelligence in Online education are the cost, mechanics and complexity behind the organization of data for the various implementations of artificial intelligent features. Furthermore, these features are not perfect. There is much research being done. Due to the limitations of a machine implementing human type thought and action when making decisions or finding information is challenging to replicate. Human thought and decision-making is not only based on knowledge, but also heuristics, bias and emotions making it quite difficult to replicate with a machine.

In the video, “How to Grow a Mind: Statistics, Structure and Abstraction”, Joshua Tenenbaum discusses how to bring machines closer to human learning and offers possible computational learning models.

“I will argue that people’s everyday inductive leaps can be understood in terms of (approximations to) probabilistic inference over generative models of the world. These models can have rich latent structure based on abstract knowledge representations, what cognitive psychologists have sometimes called “intuitive theories”, “mental models”, or “schemas”. They also typically have a hierarchical structure supporting inference at multiple levels, or “learning to learn”, where abstract knowledge may itself be learned from experience at the same time as it guides more specific generalizations from sparse data”(Tenenbaum, 2010)

The Knowledge Base

Knowledge bases are used to create expert systems. They are the “smart” in the databases used in “intelligent” systems. Knowledge bases not only contain data, but also rules, which are used with logic programming to dictate how to use, apply and interpret the data. Many learning systems implement rules in their functionality. These rules are sometimes automatically included and sometimes they are generic allowing the user to customize them for their particular use. This use of the data with rules exhibits the learning in machine learning. Having the machine learn as it interacts with real-world scenarios and users, facilitates and produces human-like processing and interactions to produce an outcome. This outcome at times may be produced at a quicker rate and at times more accurate than its human counterpart, but, at other times, its accuracy is subject to the inherent limitations of machines.

Expert Systems

eMycin is an early expert system that was used in the medical field. It helps doctors arrive at a correct diagnosis and dosage for medical patients. (van Melle, 1981). The construct of this system is the foundation for expert systems today and the foundation for the artificial intelligence components in many of today’s educational online applications.

In the diagram below, a system designer creates a knowledge base which contains expert information for the system focus, in this example, medical data and heuristics for the medical field. With logic programming and a neural network driven knowledge base, the data and rules are used to produce results.

For the Emycin expert system, the consultation driver is a logic program that traverses through the data using the rules that form the neural network and arrives at a suggested diagnosis and dosage for a medical patient.

(Van Melle, Shortliffe & Buchanan, 1981)

One of the earlier expert-based tutoring system is Goudin. In 1979, Clancey completed a tutoring system interface to Emycin, called Goudin. Within this system, Goudin interfaces with the Emycin rules to provide students with an interactive intelligent tutoring system. (Clancey, 1988).

Machine Learning: Neural and Semantic Nets

Artificial neural networks and semantic nets are some of the mechanics of machine learning implementations. Programming languages such as Lisp created by John McCarthy in 1958 and Prolog created by Alain Colmerauer and Philippe Roussel in 1972 (Computer History Museum, 2015) allowed for the logic programming needed for traversal and sometimes the creation of the neural networks to recognize the relationships between the data and their rules. Semantic nets represent knowledge in tree-like patterns connecting nodes and arcs based on these rules.

Here is an example of a semantic net for a medical dictionary.

(Cimino, 2000)

The Semantic Web

With the abundance of information on the Internet, the semantic net has been expanded to the semantic web, as well as, data mining. The Semantic Web is about making the Web more understandable by machines (Heflin & Hendler, 2001). It is also about building an appropriate infrastructure for intelligent agents to run around the Web performing complex actions for their users (Hendler, 2001). (Devedzic, 2004)

Vladan Devedžić (2004) surveys some of the important issues related to Web Intelligence and discusses their implications for Web-based teaching and learning in his paper, “Web Intelligence and Artificial Intelligence in Education”. “The scope of WI as a research field, as proposed by Zhong et al. (2002), encompasses Web information systems environments and foundations, ontological engineering, human-media interaction, Web information management, Web information retrieval, Web agents, Web mining and farming, and emerging Web-based applications. It also aims at deepening the understanding of computational, logical, cognitive, physical, and social foundations as well as the enabling technologies for developing and applying Web-based intelligence and autonomous agents systems (Liu et al., 2003).”

Natural Language Processing

Noam Chomsky

Another aspect to Artificial Intelligence is natural language processing. Avran Noam Chomsky, a professor at Massachusetts of Technology (M.I.T.) is a founder and proponent of natural language processing. Natural language processing allows for systems to process and understand data or input that is delivered in a natural language format. Semantic nets, lexicon and statistical analyzers are used in natural language processing to enable the machine to have an approximate understanding of the input and then based on the focus of the system will have rules for the output. Some learning management systems and automatic essay grading systems use this strategy. Here is a video of an interview withNoam Chomsky on Statistical Natural Language Processing (2014)

Examples of Artificial Intelligence in Online Education Systems

Artificial Intelligence can be used in various ways to enhance educational systems. One way is with the use of AI computational models and game theory as seen in games and simulations. Another is in the production of learning environments that mimic a learning strategy for teachers and learners such as collaborative, project-based or learning management systems and MOOCs. Another is the collaboration of artificial neural networks and knowledge bases with learners as seen in expert and intelligent tutoring systems.

International Artificial Intelligence Education Society (IAIED)

The International Artificial Intelligence Education Society (IAIED) is an international society bringing together computer science, education and philosophy. “AIED is an interdisciplinary community at the frontiers of the fields of computer science, education and psychology. It promotes rigorous research and development of interactive and adaptive learning environments for learners of all ages, across all domains. The society brings together a community of members in the field through the organization of Conferences, a Journal, and other activities of interest.”(International Artificial Intelligence in Education Society, 2015)

Games and Simulations

Games and simulations are used in a variety of ways in education. Game theory and artificial intelligence have similar roots. Game theory is not only used in games, but in simulations and other real-world scenarios seen in today’s educational applications. According to the Encyclopedia for Business, “….game theory has been applied to participants in parlor games and game theory scenarios, it has also been applied to a variety of real-life situations—to general human and institutional behavior—which has led the social sciences to work with game theoretic models. Numerous fields including biology, computer science, economics, politics, psychology, mathematics, philosophy, and sociology use game theory models.”

Collaborative Systems

One of the contributions of AIED is a case study collaborative system proposed by Rosalind and Self (2004) that enables collaboration between two or more learners at a distance on a case study activity. Some more recent developments emphasizing collaborative systems are Scholar (2015) a product of the University of Illinois and Common Ground andPeerStudio (Kulkarni et al, 2014) where both systems implement peer collaboration, interaction and feedback as the basis for the learning outcomes.


MOOC stands for Massive Open Online Course. cMOOC stands for Connectivist Massive Open Online Course. xMOOCs are MOOCs based on traditional university courses and sometimes offers college credit.

Udacity, edX, and Coursera are three major xMOOCs having originated in Artificial Intelligence Labs. Udacity is founded by Sebastian Thrun who also was the lead of Google’s self-driving car project. “Thrun announced his departure from Stanford where he’d been a research professor and the director of SAIL, the Stanford Artificial Intelligence Laboratory. Now the director of SAIL is Andrew Ng, who along with fellow Stanford machine learning and AI professor Daphne Koller, is the founder of Coursera. In March, Anant Agarwal announced that he was stepping down as the director of CSAIL, MIT’s Computer Science and Artificial Laboratory in order to become the president of MITx (now edX).“(Waters, 2012).

As stated by Nicholas Carr (2012), in “The Crisis in Higher Education” there are many artificial intelligence implementations in these xMOOCs, “It’s hardly a coincidence that Udacity, Coursera, and edX are all led by computer scientists. To fulfill their grand promise—making college at once cheaper and better—MOOCs will need to exploit the latest breakthroughs in large-scale data processing and machine learning, which enable computers to adjust to the tasks at hand. Delivering a complex class to thousands of people simultaneously demands a high degree of automation. Many of the labor-intensive tasks traditionally performed by professors and teaching assistants—grading tests, tutoring, moderating discussions—have to be done by computers. Advanced analytical software is also required to parse the enormous amounts of information about student behavior collected during the classes. By using algorithms to spot patterns in the data, programmers hope to gain insights into learning styles and teaching strategies, which can then be used to refine the technology further. Such artificial-intelligence techniques will, the MOOC pioneers believe, bring higher education out of the industrial era and into the digital age.”(Carr, 2012)

Intelligent Tutoring Systems

Marvin Minsky

Marvin Minsky and his colleagues are noted as some of founders of intelligent tutoring systems,” …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)

Intelligent tutoring systems are currently based on expert systems and semantic nets or semantic web-based systems. Some examples are Artimat, SmartTutor and ZOSMAT.

Artimat is an artificial intelligence-based distance education system created to help 10th grade students solve math problems. (Nabiyev, 2013). SmartTutor is an intelligent tutoring system implemented for distance learning in Hong Kong. (Chenuga et al, 2003).

Another AIED initiative is for semantic web-based educational systems (SWBES). “The main idea is to incorporate semantic web resources to the design of AIED systems aiming to update their architectures to provide more adaptability, robustness and richer learning environments. (Bittencourt et al, 2009). ZOSMAT is an example of a Web-based intelligent tutoring system for teaching–learning process. (Keles, et al, 2009).

The Seven Affordances

(Kalantzis & Cope)

The seven affordances are Ubiquitous Learning, Active Knowledge Making, Mutlimodal Meanings, Recursive Feedback, Collaborative Intelligence, and Metacognition (Kalantzis & Cope). Artificial strategies, implementations and inclusion in online educational systems can offer expressions of these affordances. Analyzing a learning system according to these seven affordances allows for its comprehensive expression under the framework of New Learning as defined by Dr. Cope and Mary Kalantzis.

Ubiquitous learning allows for learning that “extends beyond the walls of the classroom and the cells of the timetable. Learning that breaks out of these spatial and temporal confinements, should be as good as, or even better than, the best traditional classroom learning. It should also produce habits of mind appropriate to our times, producing lifelong learners, able to learn and to share knowledge throughout their lives, in all contexts, and grounded in those contexts” (Kalantzis & Cope) Artificial inteligence enhanced online educational systems such as games, MOOCs and intelligence tutoing systems can allow for self-directed, quality, ubiquitous modes of learning.

Active Knowledge Making allows for active knowledge production. Artifiical intelligent systems at their core rely on knowledge bases in usage and production both by learners and machines.

Online educational systems allow for various digital media to be used in knowledge representation and production offering multimodal meanings to the artifacts that are used and produced. Artificial intelligence strategies in semantic and web net, data mining and big data allows for the inclusion of larger amounts of digital media enriching the multimodal offerings.

Recursive Feedback can be found in the “new generation of assessment systems: Including continuous machine-mediated human assessment from multiple perspectives (peers, self, teacher, parents, invited experts etc.), and machine feedback (selected and supply response assessments, natural language processing).” (Kalantzis & Cole) many of which are made possible due to the artifical intelligence enhancements to assessment products currently being used and researched.

Artificial intelligence enhanced online systems such as collaborative systems like Peer Studio and Scholar illustrate the affordance of collaborative intelligence where the opportunity to build “skills of collaboration and negotiation necessary for complex, diverse world. It focuses on learning as social activity rather than learning as individual memory.” (Kalntzis & Cope).

Artificial intelligence can be added to online educational systems to add to the complexity of the interaction and thought provoking experience of the learner thereby promoting the use of metacognition.

Differentiated Learning allows for various forms of learning based on various factors and learning styles. This affordance can be represented not only in online educational systems through the use of various resources and avenues included, but can be enhanced with the use of artifcial intelligence computational models and strategies adding to the richness of possibilities and offerings.

Critical Reflections

Some of the challenges in using Artificial Intelligence in online education is knowing how, where and when to use them effectively, as well as, integrating the strategies with accuracy, a small learning curve and user- friendliness. Some learning theories are enhanced through the use of AI, some are not fully realized and with the use of AI can be improved. Artificial Intelligence can help in creating and finding resources easier, enhance interactivity, promote collaboration and aid in tutoring. The mechanics behind the organization and use of data, relations and rules for these features are not perfect and there is much research being done. Furthemore, the technologies can be expensive to design and implement.


With the rapid growth in online learning, artificial intelligence techniques and strategies are used to enhance the online classroom and promote an advanced learning environment. From sophisticated games and simulations to collaborative platforms and Massive Open Online Course (MOOCs) to intelligent tutoring systems, sophisticated and complex learning environments will be the norm of the future. Furthermore, with the inclusion of the seven affordances, the artificial intelligence enhanced online educational system will illustrate New Learning.


Bittencourt, I., Evandro, C., Marlos, S., Elvys, S. (2009). A computational model for developing semantic web-based educational systems. Knowledge-Based Systems, Volume 22, Issue 4, May 2009, Pages 302–315

Carr, N., (2012). The Crisis in Higher Education, MIT Review. Retrieved from

Chenuga, B., Huib, L., Zhangc, J., Yiub, S.m., (2003). SmartTutor: An intelligent tutoring system in web-based adult education. Journal of Systems and Software, Volume 68, Issue 1, 15 October 2003, Pages 11-25.

Cimino, J. (2000). From Data to Knowledge through Concept-oriented Terminologies Experience with the Medical Entities Dictionary. Journal of the Medical Informatics Association. Oxford Journals. Retrieved from

Clancy, W. (1987). “Knowledge-based Tutoring: The GUIDON Program”. MIT Press Series in Artificial Intelligence, July 29, 1987.

Computer History Museum, (2015). John McCarthy. Retrieved from,McCarthy/.

Devedzic, V. (2004) Education and the Semantic Web. International Journal of Artificial Intelligence in Education 14 (2004) 39-65 IOS Press. Retrieved from

Devedzic, V. (2004). Web intelligence and artificial intelligence in education. Journal of Educational Technology & Society, 7(4) Retrieved from

Encyclopedia for Business, (2015), Retrieved from

International Artificial Intelligence in Education Society, (2015). Retrieved from

Kalntzis, M., Cole, W., (2015) New Learning Retreived from

Keles, R., Ocak, R., Keles, A., Gulcu, A., (2009). ZOSMAT: Web-based intelligent tutoring system for teaching–learning process, Expert Systems with Applications, Volume 36, Issue 2, Part 1, March 2009, Pages 1229–1239

Kulkarni, C., Socher, R., Bernstein, M.S., and Klemmer, S.R. (2015) PeerStudio: Rapid Peer Feedback Emphasizes Revision and Improves Performance, Proceedings of the Second (2015) ACM Conference on Learning @ Scale, Pages 75-84 ISBN: 978-1-4503-3411-2 doi>10.1145/2724660.2724670

McCarthy, J. (2007). What is Artificial Intelligence? Retrieved from

McCarthy, J. et al (1955) A Proposal For The Dartmouth Summer Research Project On Artificial Intelligence. Retrieved from

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.

Morrison, D. (2013) The Ultimate Student Guide to xMOOCs and cMOOCs Retrieved from

Nabiyev, V. V., Karal, H. H., Arslan, S. S., Erumit, A. K., & Cebi, A. A. (2013). An Artificial Intelligence-Based Distance Education System: Artimat. Turkish Online Journal Of Distance Education (Tojde), 14(2), 81-98.

Tenenbaum, J., (2010). “How to Grow a Mind: Statistics, Structure and Abstraction”, Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, MIT, published: Jan. 12, 2011, recorded: December 2010. Retrieved from

Van Melle, W., Shortliffe, E.H., Buchanan, B.G.,(1981) EMYCIN: A Knowledge Engineer’s Tool for Constructing Rule-Based Expert Systems. Retrieved from

Waters, A. (2012). Top Ed-Tech Trends of 2012: Automation and Artificial Intelligence. Retrieved from

Womeundi, (2014). Noam Chomsky on Statistical Natural Language Processing. Retrieved from


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s