|
Title:
|
KNOWLEDGE UTILITY OPTIMIZATION FOR TIME-DEPENDENT LEARNING |
|
Author(s):
|
Yacine Atif , Rachid Benlamri , Jawad Berri |
|
ISBN:
|
972-8924-05-4 |
|
Editors:
|
Kinshuk, Demetrios Sampson and Pedro Isaías |
|
Year:
|
2005 |
|
Edition:
|
Single |
|
Keywords:
|
Knowledge Management, Time-Dependent Learning, Adaptive Learning, Multimedia Support for Learning, Learning Objects, Hypermedia Courseware. |
|
Type:
|
Short Paper |
|
First Page:
|
433 |
|
Last Page:
|
439 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
Humans have the ability to change and adjust their learning strategies according to different learning situations [1]. Although such ability has been largely considered in educational studies, a specific attention must be given to promote this ability in providing the appropriate contextual learning environments. The present work introduces a framework that takes contextual factors into account by facilitating learning resources to stimulate the learning capabilities of the learner. Such factors include the cognitive time relative to the typical length of a knowledge volume embedded in a learning unit. Another factor includes the effective knowledge distilled from a learning unit. This factor describes the utility value or semantic density of a knowledge resource. The framework described in this paper provides a realistic model for human cognitive processes, which are mapped into knowledge construction algorithms. In one proposed algorithm, called informative algorithm, knowledge construction is pre-planned to optimize knowledge quality based on the allocated learning time. This approach lends itself to an automated knowledge construction process, as knowledge utility is statically determined prior to exposing any material to the learner. In an alternative greedy- approach, labeled in this paper a myopic, the learner participates directly in the utility optimization process throughout a progressive knowledge construction session. This approach is greedy as it is shortsighted in the sense that it takes decisions on the basis of information at hand without worrying about the effect these decisions may have in the future. A performance evaluation of both approaches is presented in this paper. The results show interesting performance tradeoffs in scaling up knowledge quality in situations where learning time is a critical resource. |
|
|
|
|
|
|
