Research On Knowledge Reorganization Method For Emergency Quick Response

In the process of emergency quick response, decision-makers need to get the decision-making knowledge as soon as possible for quick and effective command-assignment. This knowledge usually consists in the government documents related to emergencies. According to granularity of knowledge, it has different levels, which may be the whole documents on macro-level, the physical structural on middle level, or the logical structural cells that has semantic among the paragraphs on micro-level. At critical situation, decision-makers should get the knowledge pieces quickly. But emergency documents are various, and the needed knowledge is not organized according to the order or flow of emergencies. So, it is very difficult to get the knowledge pieces for the decision-makers. For solving the problems above, three knowledge Integration methods are proposed in this thesis.The mainly contents of this paper are as followings:First, the emergency documents' system is discussed, including resources of the knowledge for emergency quick response, the meaning and classification of the knowledge for emergency quick response, the classification of emergency documents, the relation between emergency plans and emergency laws, and the analysis of emergency plans' text structural.Second, several methods of knowledge integration based on different levels are proposed, according to the classification of emergency documents, the text physical structural, and the text logical structural. Because of the difficulties in obtaining the knowledge pieces, the process of building the text logical structural is discussed in this thesis. Driven by the user's knowledge demand, a question-tree is generated, and also the text logical structural is built.Third, creating the frame of knowledge navigation and indexing, and a system for knowledge quick response is designed. It can not only provide the knowledge, like a whole document, a text physical knowledge cell, but also provide decision-makers the knowledge pieces, which have semantic characters among the related paragraphs.Fourth, we set up five factors to evaluate the results of knowledge integration, and implement it by using method of fuzzy integrated evaluation in this thesis.