DATA MINING MODEL BASED FOR DETECTION FRAUDULENT OF BEHAVIOUR IN WATER CONSUMPTION

Abstract

Fraudulent behavior in drinking water consumption is a significant problem
facing water supplying companies and agencies. This behavior results in a massive
loss of income and forms the highest percentage of non-technical loss. Finding
efficient measurements for detecting fraudulent activities has been an active research
area in recent years. Intelligent data mining techniques can help water supplying
companies to detect these fraudulent activities to reduce such losses. This research
explores the use of two classification techniques (SVM and KNN) to detect suspicious
fraud water customers. The main motivation of this research is to assist Yarmouk
Water Company (YWC) in Irbid city of Jordan to overcome its profit loss. The SVM
based approach uses customer load profile attributes to expose abnormal behavior that
is known to be correlated with non-technical loss activities. The data has been
collected from the historical data of the company billing system. The accuracy of the
generated model hit a rate of over 74% which is better than the current manual
prediction procedures taken by the YWC. To deploy the model, a decision tool has
been built using the generated model. The system will help the company to predict
suspicious water customers to be inspected on site.