Association Rules on Traffic Accident: Case Of Ankara


Creative Commons License

Soylemez I., DOĞAN A., ÖZCAN U.

EGE ACADEMIC REVIEW, cilt.16, ss.11-20, 2016 (ESCI İndekslerine Giren Dergi) identifier

  • Cilt numarası: 16
  • Basım Tarihi: 2016
  • Doi Numarası: 10.21121/eab.2016ozel24423
  • Dergi Adı: EGE ACADEMIC REVIEW
  • Sayfa Sayıları: ss.11-20

Özet

In this study, association rules analysis of the data mining techniques are used for data of traffic accidents in 2010 and some rules are obtained. With this rules, what is the possibility of accident which resulted anybody injured for "different weather conditions (snowy, rainy etc.)", "where the accidents occurred (street, road etc.)" and "way situations (separated road or not)". Different algorithms are used to analyze the association rules. Apriori algorithm is selected for this study and SPSS Clementine 12.0 is used for this algorithm. Firstly, frequency of items are found. Then, items are grouped. In this study, data preprocessing is done and missing values are filled or rejected. In the second phase, outliers are rejected and data type is converted type of 1-0 (binary). In the third phase, Apriori algorithm is applied and results are evaluated.