Data Mining Technology Applied Research In The Analysis Of Fire Incidents

In recent years, though the number of fire accident has decreased gradually, there were great losses every year in China. In the past 4 years, the number of fire accident decreased from 235,000 in 2005 to 74,000 in 2009 nationwide. Meanwhile, the accumulation of fire accident record provided plenty of first-hand information for the research of cause, process and control method of the accidents. Finding out the potential law of fire accidents is the key point to control the fire accident from the source, and the foundation of which is analysis of plenty accident records. How to manage and analyze the data? How to find the connotative law of accident? The advent of data mining technology provides an effective solution to these problems.The thesis is written under this kind of background. The works of thesis is sorting out and studying the fire accident data by combining data mining technology with accident analyzing theory, and tentatively mining the data with relative mining tools.The study is mainly about the application of mining in fire accident data. After a brief introduction of background of this subject, the thesis presented the concept, development and method of data mining. Also, the analysis model and common Algorithm is explained in chapter 4.In respect of accident data, for the deficiency of accident database and simplicity of statistic method, the study adopted Microsoft Office Access database system to reconstruct the database and realize the sorting, refining and converting of current data. The content of database is Building the database of accident data and statistical data separately by combining the theory of data warehouse and the accident analyzing technology.The study adopt Microsoft SQL Server Business intelligent component to mine the refining data, Using cluster, relative rule, time serial and decision tree method mined the accident and the statistical data from different aspect. The result of the data mining of accident explained the law of accident and the inter connection between the accident attribute to some extent.The result of this study is expected to provide a reference for the application of accident analysis, and have academic value and realistic meaning to fire accident data processing and information analysis.