Detection of Interesting Traffic Accident Patterns by Association Rule Mining

Abstract

In recent years, the accident rate related to traffic is high. Analyzing the crash data and extracting useful information from it can help in taking respective measures to decrease this rate or prevent the crash from happening. Related research has been done in the past which involved proposing various measures and algorithms to obtain interesting crash patterns from the crash records. The main problem is that large numbers of patterns were produced and vast number of these patterns would be obvious or not interesting. A deeper analysis of the data is required in order to get the interesting patterns. In order to overcome this situation, we have proposed a new approach to detect the most associated sequential patterns in the crash data. We also make use of the technique, “Association Rule Mining” to mine interesting traffic accident patterns from the crash records. The main goal of this research is to detect the most associated sequential patterns (MASP) and mine patterns within the data sets generated by MASP using a modified FP-growth approach in regular association rule mining. We have designed and implemented data structures for efficient implementation of algorithms. The results extracted can be further queried for pattern analysis to get a deeper understanding. Efficient memory management is one of the main objectives during the implementation of the algorithms. Linked list based tree structures have been used for searching the patterns. The results obtained seemed to be very promising and the detected MASPs contained most of the attributes which gave a deeper insight into the crash data and the patterns were found to be very interesting. A prototype application is developed in C# .NET.