For over 40 years, researchers have been trying to develop ‘intelligent’ computer technology to improve the delivery of instruction to learners and help them learn faster and more efficiently. The general feeling among educators and researchers alike is that these efforts have not been successful, at least in terms of their widespread adoption in education.

Intelligent tutoring systems (ITS) differ from the previous generation of computer-aided instructional tools because they try to model the domain being taught and the student’s likely mastery of its content. An adaptive and completely personalised path through the content is dynamically and continually constructed by the system, based on continuous assessment and feedback.

A large number of studies suggest that students taught using ITS in well-defined domains, such as mathematics, can learn to the same level of mastery as those taught by traditional classroom teaching and even one-to-one human tutoring. In addition, the time to reach that level of competency is generally shorter when compared with more traditional methods.