<![CDATA[Road damage due to ODOL (Over Dimension Over Loading) increases the road maintenance budget significantly, namely an average of IDR 43.45 trillion per year. In addition, many accidents involving ODOL trucks or overloading and dimensions have occurred. The level of violations caused by ODOL vehicles is still high, so technology is needed that is able to manage data and serve as a reference to find out the hidden approaches in the data set, as well as analyze the grouping between data and attributes to facilitate decision making and policy making. This study applies the CRISP-DM methodology using a decision tree model with the C4.5 algorithm. The purpose of this research is to classify trends in freight transport violations based on violation data in the UPPKB. The research data is primary data obtained from the Directorate of Road Transportation Infrastructure of the Ministry of Transportation through the online weighbridge system (JTO). The expected result of this research is to be able to find out the pattern of classification trends for freight vehicle disturbances based on the results of the C.45 algorithm decision tree, so that the research results can be used as a reference in making decisions and making policies. The results of this study indicate that the accuracy performance in data mining tests for the classification of trends in freight vehicle disturbances with 10 fold cross validation linear sampling produces an accuracy of 86.31% +/- 1.23% (micro average: 86.31%), shuffled sampling produces an accuracy of 86.34% +/ - 0.67% (micro average: 86.34%) and stratified sampling produces an accuracy of 86.34% +/- 0.67% (micro average: 86.34%).]]>
