Case Retrieval Optimization Method for Traffic Accidents

Abstract:

An optimization method for traffic accident case retrieval is proposed to improve the accuracy of real time decision management under traffic accident conditions. Based on traffic accident data constitution of Numeric and enumeration type, attribute weights are determined objectively through information entropy method evaluating the data dispersion degree. On the basis of data classification, case retrieval base is established through second order clustering algorithm. And a traffic accident case retrieval system is developed to have a series of tests. Then, case retrieval tests are conducted to verify effectiveness of this optimization method by analyzing both highest similarity and case set matching degree of the retrieval conclusions.