Research Of E-Learning Personalization And Resource-Sharing Based On Ontology

With the rapid development of Internet and global requirement for life education, e-Learning which providing asynchronous education has been becoming an important application on Internet. The distributed environment of Internet can provide powerful technical supports for distance education and knowledge services on the aspect of knowledge acquirement and learning collaboration. The learning resources on Internet are growing richer and richer, but the current e-Learning application still remains some limitation, such as personalized study and knowledge sharing in the dynamic web environment. Aiming at these problems, some investigation including knowledge semantics and architecture are made in this thesis.We firstly discussed the usefulness of learning resource standards and learning objects for e-Learning environment. These standards can support personalization and resource-sharing to a certain extent, but they still have shortcomings such as their exclusive focus on property based specifications and their capability of describing subject information of learning material. On the basis of analyzing the functionality of personalization and resource-sharing, we pointed out that its underlying reason is the weak-semantic resources and the weak sharing mechanism. Thus we proposed to depict the resources semantics by ontologies and implemented the e-Learning system in the semantic web framework, thereby improving the personalization and resource-sharing level through semantic interoperability.As conceptual model, ontology can express the basic knowledge system of a domain. Based on semantic web, we investigated the need and feasibility of the ontology applying in e-Learning environment, and specified the semantic techniquesfor required levels of e-Learning domain. Then we discussed the semantic capability of RDF/S, and used RDF/S as ontology describing and TRIPLE as ontology inferring, so to realize intelligent search and aviation of e-Learning resources according to the ontology model.Based on ontology modeling, we established the e-Learning domain semantic system, defined some ontologies in formalization, and gave some examples. It included domain knowledge ontologies, learner ontologies, and course concept ontologies. Accordingly, we separated the resource contents and the relationship between them in e-Learning environment, and provided semantic basis for comprehensive resource-sharing and personalization.After the building of e-Learning domain knowledge ontologies, we proposed a semantic describing and inferring based adaptive e-Learning architecture using layered principle modeling, and analyzed functions and techniques of each level of the system. Layered framework makes the system become modular and open, and can provide better personalization sevices. Additionally, it can be applied in the semantic web framework, which can support distributed learning resources sharing and integrating by web services.At last, we designed an adaptive e-Learning system named Ont-Edu, and discussed its implementation with a Java language curriculum. Ont-Edu can provide personalized learning content and style, and also personalized instruction according to the learner characters. It bridges the gap between adaptive educational systems and distributed learning resources environment.