Verifying Waste Disposal Practice Problems of Rural Areas In Indonesia Using the Apriori Algorithm

Abstract

Verifying a set of most frequent problems is essential before introducing practical solutions using new technology, processes, and practices. This study proposes a way to verify these problem sets. The main contribution of this paper is a method to verify a set of most frequent problems in waste disposal practices previously identified through a survey questionnaire, using Google Earth visualization and the Apriori algorithm. Google Earth is used to pinpoint the geographical locations of existing waste bins, illegal landfills, and people's houses. The distance between the waste bins and the residents' houses, sites of waste disposal by burning, and sites of waste disposal by dumping are then analyzed as a combination of the problems of waste disposal practices. Support, Confidence, multiplication between Support and Confidence, and lift ratio values are then calculated to obtain a combination of the most frequent problems sets. Next, the support value in the Apriori algorithm is compared with the FP-Growth method using Rapidminer. Results obtain support and thus verify data previously obtained from the survey. For a 2-itemset problem and a minimum support value of 0.1, 33% accuracy is obtained, while a 3-itemset problem returns 99% accuracy. We show that our method is useful in verifying data previously obtained from other sources.