Using K-Means Algorithm to Investigate Community Behavior in Treating Waste toward Smart City

Abstract

Urban communities' behavior in disposing and managing waste around the house is still very concerning. This study aims to analyze and classify community behavior patterns related to waste disposal to help the Makassar City Government overcome waste problems. The methods and techniques applied are using the Waterfall model and the K-Means Algorithm. The stages of analysis and classification of behavior patterns can assist city governments in making strategic decisions. The variables determined include education, age, occupation, free time, smoking status, and questionnaire (outreach, law, knowledge, and facilities). The K-Means algorithm calculation results are 39 people in cluster 1 and 10 people in cluster 2. In cluster 1, it is known that the community tends to take care of the environment in the sub-district where the community lives. Cluster 2 has a low average yield, such as people who do not know the impact laws, inadequate sanitation facilities, and lack of proper waste management towards smart city governance. Based on testing for each method, the Cyclomatic Complexity value generated is four. Therefore, it can be concluded that the white box testing on the K-Means algorithm runs well because each test produces the same value. The government and the community have a responsibility to carry out waste management properly, make use of goods and refill facilities, know the legal impact of littering and follow the socialization held by the Makassar City Government.