Implementation of Data Analytics for the Accuracy of Service Time Prediction Models

Abstract

The performance of repair services is very important to determine the achievement of consumer confidence. The strategy that needs to be made is to pay attention to the timeliness of repairs so that no one is harmed between consumers and service providers. Tracing information on data from repair services is one effective way to determine the accuracy of computer repair time. The collection of information used comes from the Istidata Indopacific Solution Center (IISC) repair service dataset, consisting of a collection of data on the completion time of product unit repairs that are achieved and not achieved. Repair completion time is the time in accordance with the agreement between the repair service party and the consumer. Data processing is carried out by processing analytical data by utilizing the Weka Tools software with the application of classification with the J48 decision tree method which is the development of the C4.5 algorithm. The effectiveness of this method was tested using 10-fold cross validation, where from the results of the confusion matrix measurement an accuracy of 99.5% was obtained. The result states that the J48 decision tree method is effective and can be used to predict the accuracy of computer repair time.