DATA INTEGRATION TECHNOLOGY FOR MULTI-SOURCE GEO-SPATIAL INFORMATION FROM WEB FOR RAPID EMERGENCY RESPONSE

The basic geo-spatial information for disaster reduction, including environment, transportation, construction, population and so on, plays a crucial role in emergency management. With the rapid development of Internet, web has become an important channel for the rapid information acquisition. However, most of the information exists in the Deep Web that is difficult to obtain, and different sources of information can not be directly integrated due to their heterogeneity and structural diversity. Therefore, in this paper we propose a multi-source geo-spatial information integration method for disaster reduction. By constructing web focused crawler, we can realize the on-demand aggregation and classification cleaning of multi-source geo-spatial information, and implement the aggregation and duplicate removal of heterogeneous geographic entities based on semantic similarity calculation. Experiments show that this method can support the rapid integration of multi-source geo-spatial information for disaster reduction in the case of sudden disasters and provide rich geo-spatial information support for emergency management.