Traffic Accident Data Mining Based On Association Rules

With the cities’ rapid construction and technological progress,the road traffic has also been rapid developed,but there are also a growing number of traffic accidents,which made the safety of people and society facing a great threat.Although the comprehensive construction of intelligent transportation system has improved the situation of the traffic accidents that happening frequently,but it is still very serious,so it is very important to excavate and analyze the traffic accident data.Association rule technology is a classic technology method in data mining theory.It is mainly used to find out the implied relevance information among projects from a large amount of data,and help relevant personnel to make appropriate decision.In the first part of the research,association rules are mainly directed at the mining of frequent patterns.With the development of association rules theory,association rules are extended to mining association rules in many different ways.In this paper,association rules and its extension technology were used to analyze the traffic accident data after pretreatment in a city in 2015 from three aspects.Firstly,the dimension information was added to conventional FP-Growth algorithm to deal with multi-dimensional data,then the FP-Tree was constructed and all the frequent patterns can be found using the frequent pattern growth method.Secondly,the weighted FP-Growth algorithm was proposed based on the multi-dimensional algorithm.The data were analyzed by the weighted association rule theory.The weights of each influencing factor of the accident were calculated,and then more frequent patterns about accident description were excavated.Thirdly,the concept of weight vector was put forward on the basis of frequency vector,the dynamic FP-Growth algorithm was proposed by combining the weighting model,weighting support confidence vector were used to describe the association of association rules and time variation,and the most suitable time period for each association rule expression was found.Finally,the impact of various factors in the city on the traffic accident was analyzed through association rule expression.