Wyniki wyszukiwania - BazTech

Ograniczanie wyników

1 Journal of Artificial Intelligence and Soft Computing Research

1 2011

Znaleziono wyników: 1

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: adaptive educational hypermedia systems

Sortuj według:

Ogranicz wyniki do:

Learning style recognition based on an adjustable three-layer fuzzy cognitive map

Georgiou D. A., Botsios S., Mitropoulou V., Papaioannou M., Schizas C., Tsoulouhas G.

Journal of Artificial Intelligence and Soft Computing Research

2011

Vol. 1, No. 4

333--347

Identification of learning styles supports Adaptive Educational Hypermedia Systems compiling and presenting tutorials custom in cognitive characteristics of each individual learner. This work addresses the issue: identifying the learning style of students, following the Kolb’s learning cycle. To this purpose, we propose a three-layers Fuzzy Cognitive Map (FCM) in conjunction with a dynamic Hebbian rule for learning styles recognition. The form of FCMs is designed by humans who determine its weighted interconnections among concepts. But the human factor may not be as reliable as it should be. Thus, a FCM model of the system allowing the adjustment of its weights using additional learners’ characteristics such as the Learning Ability Factors. In this article, two consecutively interconnected FCM (in the form of a three layer FCM) are presented. The schema’s efficiency has been tested and compared to known results after a fine-tuning of the weights of the causal interconnections among concepts. The simulations results of training the process system verify the effectiveness, validity and advantageous characteristics of those learning techniques for FCMs. The online recognition of learning styles by using threelayer Fuzzy Cognitive Map improves the accuracy of recognition obtained using Bayesian Networks that uses quantitative measurements of learning style taken from statistical samples. This improvement is due to the fuzzy nature of qualitative characterizations (such as learning styles), and the presence of intermediate level nodes representing Learning Ability Factors. Such factors are easily recognizable characteristics of a learner to improve adjustment of weights in edges with one end in the middle-level nodes. This leads to the establishment of a more reliable model, as shown by the results given by the application to a test group of students.