摘要
城市机动车数量的增加已经导致城市交通事故的频繁发生,能否对已发生事故做出正确的分析将直接影响到能否对未来类似事故的成功避免。随着社会的发展,道路交通事故系统的复杂性也在逐渐增强,传统的分析、预防方法已呈现其局限性。现提出一种使用数据挖掘领域中的多维关联规则技术分析大量交通事故记录的方法,通过找出可能导致交通事故发生的频繁因素组合来发现某些事故发生的规律,以期为道路安全管理提供科学的决策依据。通过该方法,我们可以有效地识别和发现事故数据的新模式,且能为交通管理决策提供有效支持,该方法易于实现,便于推广。
The increase in the number of urban vehicles has led to the frequent occurrence of traffic accidents. Whether those accidents can analyzed correctly in the past will directly determine the avoidance of future ones of the similar kind. With the development of the society , tradition approach is incompetent due to complexity of the system of casualty for the analysis and prevention. A famous method is introduced in the field of data mining, called multidimensional association rule method. The method helps us to analyze the large amounts of traffic accident records. With this method, It are able to find out the underlying rules in traffic accidents through searching the combination of frequent factors that probably lead to traffic accidents. It is expected that data mining will be helpful the scientific decision making for management of traffic safety. Using this method, It can identify and find new modes of traffic accident data effectively. The method is effective in decision-making support of traffic control. It can be easily popularized.
出处
《科学技术与工程》
2008年第7期1824-1827,共4页
Science Technology and Engineering
作者简介
王云（1982-），男，江苏科技大学硕士研究生，研究方向：数据挖掘、数据仓库。E-mail：wy825290@126．com。
通信作者：苏勇（1961-），男，江苏科技大学电子信息学院副教授，硕士研究生导师，研究方向：数据挖掘、人工智能。
