Ontology Construction from Semi-Structured Text

Ontology is effective in supporting text mining, information retrieval and knowledge engineering on semantic level. However, ontology construction is pretty time-consuming; hence, some automatic or semi-automatic methods are proposed to try to solve the problem. In this paper, a model is issued to construct a medical ontology from semi-structured text after three steps: PAT-Tree based phrase extraction, document structure analysis and ontology construction based on hierarchy structure of text. This model has found some applications in medical emergency aid system in China. Practical experiment results show that this model can create a practical medical ontology from syllabus of medical school efficiently and effectively.