Implementation of address recording management using the K-Means clustering classification algorithm in Kebayoran District, DKI Jakarta

Abstract

The area is the center of problems in the administrative record management of Kebayoran District, because of its dense condition and it is difficult to determine land measurements due to the density of residential areas. The problem in Indonesia to this day is that the administrative boundaries of the kelurahan already exist, but the administrative boundaries for the Rukun Warga / Rukun Tetangga (RW / RT) do not yet exist. The local government of DKI already has a large scale map (1: 1,000) to map RW administrative boundaries. Large-scale mapping (Batas RW) is useful for accurate information on incidence of dengue fever or other diseases, thereby eliminating information bias due to the use of village boundary maps. Another benefit is the accuracy of address management for customers, for example PDAM customers, to facilitate verification of customer data with large-scale maps, especially those that only include RT / RW addresses, without mentioning street names and household numbers. The method used is data mining K-Means Clustering. By using this method, the data that has been obtained can be grouped into several clusters, where the application of the KMeans Clustering process uses Excel calculations. The processed data is divided into 3 clusters, namely: high cluster (C1), medium cluster (C2) and low cluster (C3). The iteration process of this research occurs 2 times so that an assessment is obtained in classifying the household / neighborhood unit based on the Kelurahan. The results obtained are that there is 1 neighborhood unit with the highest cluster (C1), there are 4 neighborhood units with 4 medium clusters (C2), and 5 neighborhood units with the lowest cluster (C3). This data can be input to the sub-district to disseminate information about dengue fever, health education, and for the accuracy of PDAM customer address management and others.