| Emergency decision support system requests to concentrate all the resource information, using multi-information network to get the correct emergency information and reasonable matching based on diverse environment. It can bring its effectiveness into emergency or durative opposition. Association rule mining describes the relation among all the objects in the dataset. It's very useful to adopt association rules describing the cases property relativity in decision system cases, because it could provide correct and easy used decision knowledge to decision-makers. In the thesis, the author researches on association rules based on ontology knowledge in fire emergency field.Firstly, while analyzing association rules mining in detail, the author makes a lot research on ontology knowledge expression technology oriented to association rules, establishes the ontology database in fire emergency system, and then uses the ontology to lead data mining that can get the mining process more quickly and enhance the efficiency and quality of obtaining knowledge.Secondly, after analyzing the relativity of decision data and concerning with some weakness in association rules mining, such as high cost of mode counting and inefficient I/O, the author proposes one association rules mining algorithm based on ontology. In this algorithm, candidate itemsets that are not suitable for constrained conditions are pruning, and the itemsets which user is interested are built only. The algorithm can reduce mode-counting cost and improve mining quality. In fact the author makes... |