East Java']Java Rivers Characterization Using Analytical Hierarchical Clustering on Supporting Disaster Management System

Floods are disasters that can be caused by nature or human actions. Floods come from the rivers that overflow due to the discharge exceeding the capacity of the river itself. As a result, the overflow of water will lead to the surrounding land. To avoid flooding, a flood mitigation system should be made. River characterization will be one solution of the early flood detection models. To characterize the rivers, we propose the AHC (Analytical Hierarchical Clustering) method which processes river parameters including river length, width, depth, and river water flow. The model will cluster the dataset according to the hazard level of the river. This research uses east Java rivers features provided by Ministry of Public Works and Housing of Surabaya City. The experiment result shows that the greater the number of clusters involved will cause a decrease in the accuracy of the system. The best accuracy rate is obtained as 70% when the data is divided into 2 clusters.