Road traffic accident data mining based on weighted association rules

Road traffic accident data mining can extract valuable insights from massive datasets, provide information for safety measures and decision-making, and help reduce accidents and improve overall road safety. To enhance the precision and thoroughness of road traffic accident data mining, a method based on weighted association rules has been developed. Initially, the method involves converting road traffic accident data into a distinctive one-hot encoding format. Subsequently, the transformed data is entered, and the Euclidean distance is employed to determine the similarity between datasets. The Apriori algorithm is then applied to extract frequent itemsets from the data, leading to the creation of pertinent association rules. Finally, by incorporating the concept of weights, different accident attributes are assigned varying degrees of importance, reflecting their influence on the occurrence or severity of accidents. Experimental outcomes indicate that the proposed method yields data mining results with higher coverage, lower average absolute error, reduced memory consumption, and enhanced applicability.