An Ontology Based Approach for Disaster Prediction

Natural disasters are those anomalies created by the nature in the process of the Earth. The harms they have caused to human society tend to be horrifying. Almost every country will be more or less affected by natural disasters annually. In consequence, giving a fairly accurate forecast of the disaster that may happen turns into a vital part of disaster prevention. Traditional research on disasters is mainly based on classical statistical methods or expertise judgment, which is lack of semantic-driven and intelligent reasoning. In this paper, an ontology based approach for disaster prediction was put forward. As the explicit specification of a conceptualization, ontologies can achieve a certain degree of knowledge sharing and reuse, and improve the system ability of communications, interoperability and reliability. On the basis of analyzing the influencing factors and historical evolution of disasters, a domain ontology model of natural disaster in OWL ontology language was created. The hidden influencing factors or disaster chain information can be excavated by the reasoner with domain-specific rules. As a result, we built an ontology-driven disaster prediction and information system. By performing the task of typhoon prediction based on our ontology-based system, our approach can improve the intelligence level of the disaster in its management and forecast.