Association Rule Mining for Causes of Railway Traffic Accidents Based on Improved Apriori Algorithm

LIU Zhaohui,HE Shiwei
DOI: https://doi.org/10.16668/j.cnki.issn.1003-1421.2023.04.17
2023-01-01
Abstract:Many factors can cause accidents in complex railway transportation systems, and they are correlated to some extent. Analyzing the causes and correlation characteristics of accidents is of great significance for accident prevention.According to the hierarchical analysis of accident factors at four levels, i. e., human factor level, equipment level,environmental level, and management level, the same or similar conditions were integrated, and a Python program was developed to capture the causative factors in all accident investigation reports in a recent three-year period. First, the causative factors were disassembled into relevant causative keywords, and regular expressions were used to grab them in specified paragraphs. Then, the causative factors of all accidents were obtained through keyword grabbing. Finally, 90causative factors of railway accidents were extracted. On this basis, the Apriori algorithm was improved by compression and hash table technology to mine the association rules of the causes of railway traffic accidents, and 159 association rules were found. Upon the analysis of the characteristics of high support and high confidence of accident causes, key factors, key causative associations, and causative laws of accidents were discovered, which helps to improve the level of accident prevention and safety management.
What problem does this paper attempt to address?