Aturan Asosiasi Dengan Standar Storet Pada Model Prediksi Parameter Pendukung Uji Kualitas Air Baku

Abstract

Raw Water (Air Baku) laboratory analysis is testing physical, chemical and bacteriological characteristicsof water to ensure that water supply is clean, safe and ready for drinking water quality. Analyzing raw water quality in laboratorium needs more time. The proposed system could shorten the laboratory processing time by analyzing daily water production log. Association ruleinthe proposed system was used to generate relation model of water characteristicsfrom the data log provided by local government owned water utilities (PDAM, Perusahaan Daerah Air Minum). The data was transformed first from numerical data into categorical data using STOrage and RETrieval Data Warehouse (STORET)standard.Generated model needs to be simplified because some prediction rules could have the same interpretation. The generated parameter prediction modelwas sufficient to be used as the supporting data for any local policy made related to water supply and sanitationwithout additional costs from standard lab testing of water quality. Some water quality values of chemical characteristics need lab testing. Given the missing values of several chemical characteristics, the generated parameter prediction model still could give better accuracy of 80%-95%. Since PDAM staffmanually validated the generated model, the experiments used small data set.

Keywords

• water quality prediction
• association rule
• storet