Research On The Knowledge Organization Model And The Application For E-learning

Digital educational resources are the core of e-learning. Educational resources have exerted great cognitive stress and overload upon learners of wide scope because of explosive increase of information, which are cornered in a dilemma:on the one hand, there are mass information resources, while one the other hand, these resources do not meet individual demands. Therefore, it is highly suggestive that the traditional way of organization and service of information resources should develop towards a more concise and precise knowledge service. Knowledge service, based on the object of knowledge and on the means of organization strategies like the description, categorization and relation of educational resources, equips learners with the individualized services of the resource retrieval, navigation and push, which has become a hot topic in studies on the effective application of educational resources in the field of e-learning.Digital education resources, characteristic of knowledge ability, logicality and expertise, are resources that associate classification of disciplines. Traditional method of resource organization, such as thesaurus, taxonomy and metadata description etc., can not realize the internal relations of mass educational resources to make effective organization and management about educational resources, which results in inefficiency and ineffectiveness of educational resource application. On the other hand, digital education resources cover various departments of disciplines, different bodies of knowledge and complicated semantic relations, bringing new requests for discipline-oriented knowledge understanding and semantic-interconnected knowledge organization method. To be aimed at problems existing in the application of educational resources, the present dissertation studies disciplinary ontology-based knowledge organization model. By virtue of educational technology standards (ETS) and natural language processing (NLP) technology, it probes into auto-retrieval of education resource descriptors and discipline auto-classification. It also explores application strategies and methods of the knowledge organization model. The research work mainly covers the following aspects of problems:(1) Knowledge organization model of educational resources. The dissertation constructs a disciplinary ontology-based organization model of educational resources, including resource semantic features, metadata features and linked characters. The model is made up of physical layer, logical layer, ontology layer and user interface layer. It solves the problem of heterogeneity in knowledge representation of educational resources, shielding the unconformity of resource entities through logical description and organization, so as to provide effective support for users’visit and search efficiency. Besides, the dissertation takes the discipline of education technology for an example, consulting relevant classification in Chinese Library Classification, realizing the knowledge organization classification of educational resources in the discipline of education technology, verifying and analyzing the effectiveness of the model through instance analysis.(2) Discipline-oriented auto-extraction method of educational resource of the metadata. The dissertation classifies the descriptive information if educational resources into descriptive metadata and metadata of semantic content. It proposes a discipline-oriented metadata extraction of education resources which combines rules and statistic model. The method auto-extracts descriptive metadata with a combination of features like keyword, location, file structure information etc. It auto-extracts semantic content metadata with a combination of disciplinary descriptor, disciplinary knowledge classification etc. The dissertation designs and realizes a metadata extraction system of educational resources whose feasibility is tested with more than 2000 journal articles in the discipline of education technology.(3) Disciplinary feature-oriented auto-classification method of educational resources. The dissertation puts forward a discipline-oriented, keyword feature-based auto-classification method of educational resources. It constructs a discipline-oriented topic classification model of educational resources combining knowledge classification system in the discipline of education technology. It builds an education technology thesaurus of descriptors, with kernel textbook and core journal articles in education technology as knowledge source. It suggests a discipline-oriented, keyword-based discipline classification system, explores the discipline of education technology-oriented auto-classification system of educational resources by constructing a discipline-oriented theme feature extraction method and a subject discipline-based text classification method. Results in experiments verify the efficiency of the discipline-oriented auto-classification method of educational resources.(4) Knowledge organization-based education resource navigation prototype system and its application. Knowledge organization model of educational resources is applied in the organization, navigation and retrieval of digital education resources and a knowledge organization model-based digital education resource navigation prototype system is constructed. It takes the discipline of education technology as application background, using education resource knowledge organization model to realize the knowledge organization in websites of education technology and realize the category navigation for web-based educational resources.The features and originalities of the dissertation lie in that:first, it constructs a disciplinary ontology-based organization model of educational resources, realizes the knowledge organization classification of educational resources in the discipline of education technology, verifies and analyzes the effectiveness of the model through instance analysis; second, based on disciplinary knowledge classification features, it proposes a discipline-oriented metadata extraction of education resources which combines rules and statistic model; third, it proposes a disciplinary feature-oriented auto-classification method of educational resources. The research results are of both theoretical and practical certain significance to the knowledge service system under construction.