Extracting Meta-knowledge from Multi-source Knowledge base with Concept Segmentation Method

The paper proposes a concept segmentation method to extract meta-knowledge from the multi-source knowledge base. We improve the traditional structure-based extracting method by using the concept hierarchical partition. The concept and concept relationship can be described with ontology model, which can discover the semantic relationship between concepts. Then a self-learning of meta-knowledge model is set up which can optimize the meta-knowledge description. Finally an empirical study is carried out by implementing the meta-knowledge extraction process from multi-source knowledge bass for educational resources.