Irshad:

Here are linkf from similar RG questions that can be useful:

https://www.researchgate.net/post/How_we_can_employ_adaptive_learning_strategy_in_our_classrooms2

https://www.researchgate.net/topic/intelligent_tutoring_systems

Best regards,

Debra

It's a little more complicated than which one is the best and many of the ones that have been shown to be effective aren't necessarily widely available. It's important to consider which ones are effective for the specified domain and context.

Here are some papers that may help you start to sort through the literature that you are finding.

VanLehn, K. (2006). The Behavior of Tutoring Systems. Int. J. Artif. Intell. Ed., 16(3), 227–265.

VanLehn, K. (2011). The Relative Effectiveness of Human Tutoring, Intelligent Tutoring Systems, and Other Tutoring Systems. Educational Psychologist, 46(4), 197–221.

Martin, B., Mitrovic, A., Koedinger, K. R., & Mathan, S. (2010). Evaluating and improving adaptive educational systems with learning curves. User Modeling and User-Adapted Interaction (UMUAI), 21(3), 249–283. http://doi.org/10.1007/s11257-010-9084-2

Mitrovic, A., Koedinger, K. R., & Martin, B. (2003). A Comparative Analysis of Cognitive Tutoring and Constraint-Based Modelling. In P. Brusilovsky, A. T. Corbett, & F. de Rosis (Eds.), 9th International Conference on User Modeling (UM) (pp. 313–322). Johnstown, PA, USA: Springer-Verlag. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.101.1952

The International Journal of Artificial Intelligence in Education (IJAIED), User Modelling and User Adapted Interaction (UMUAI), and IEEE Transactions on Learning Technologies (TLT) will be good places to start your investigation of this area in a less overwhelming way.

Dear Carrie,

Thank you so much for the reply. The context I am referring is to develop a system which can efficiently evaluate a student's knowledge in the initial stage and monitor his/her learning in latter stages and give him/her feedback until he/she masters a particular topic/subject. 

My first favorite theory was Knowledge Space Theory(KST) (This also used by ALEKS). Later I came across this paper : http://caaconference.co.uk/wp-content/uploads/Sitthisak_caa2013_submission.pdf .Which basically integrates KST and Competency model (COMBA)  to come up with a new model (ComKos) which is the best of both.

Now I am confused which one to use? I hope you understand my need more precisely now. Can you tell me which one should I use? 

Note: To my knowledge at least I haven't seen people using ComKos model in real life applications (according to my search).

Thanking you in advance.

You might have a look at

Kulik, J.A., & Fletcher, J.D. (2016) Effectiveness of Intelligent Tutoring Systems:  A Meta-analytic Review. Review of Educational Research, 86,42-78.

Probably the most effective ITS is DARPA's Digital Tutor --

Fletcher, J.D. (2014 ). Digital Tutoring in Information Systems Technology for Veterans: Data Report (D-5336). Alexandria, VA: Institute for Defense Analyses.

Fletcher, J.D., & Morrison, J.E. (2014). Accelerating Development of Expertise: A Digital Tutor for Navy Technical Training (D-5358). Alexandria, VA: Institute for Defense Analyses.

But that concerns very specialized (Information Systems Technology training for Navy technicians) and (the last I saw) really expensive to use.  The folks who produced it are now working with the Gates Foundation on a pre-algebra tutor, which I  understand is similarly effective.  No third party evaluation of it as yet..

Forgot to say that the Veterans report is currently in press (Educational Technology Research and Development).

I can't answer which one you should use. It depends far too much on the specific contextual details and the trade-offs you are willing to make.

With respect to mastery learning, Bayesian knowledge tracing is a commonly used approach. Koedinger and Aleven have done a fair bit of work in this area. You might want to look into the availability of those approaches.

A few years ago, only computer programmers were capable of building intelligent tutors. It involved hundreds of hours of work from several computer programmers to be able to build a tutor that is capable of training students in a very small domain of knowledge. This was a major shortcoming: programmers did not have sufficient subject knowledge to think carefully about how students learn, whereas teachers with pedagogical knowledge did not have the skill to build intelligent tutors. Recently, the trend is changing. With the improvement of graphical user interfaces and basic programming knowledge, teachers are now capable of building intelligent tutors that can cater to their particular student group's learning preferences.

For example, see cognitive tutor authoring tools (CTAT) from Carnegie Mellon University. This tool helps teachers build computer tutors with little or no programming knowledge. Here are some of our articles in which we built tutors using CTAT:

Article A Personalized Study Method for Learning University Physics

Preprint A Personalized Computer Based Tutor for Vector Algebra

Hope this helps.

Best

Dr. V. Aravind

Conference Paper Intelligent Tutoring Systems for Collaborative Learning: Enh...

Article Intelligent Tutoring Systems: an overview

Poster An Empirical Evaluation of Adaptation based on Learning Styl...

Following.