Penerapan Metode SMOTE dan K-Nearest Neighbor untuk Klasifikasi Tingkat Kepuasan Masyarakat terhadap Penyiaran Radio Most FM Medan

Abstract

Radio Most FM is one of the leading radio stations in Medan that continues to improve the quality of its broadcast programs to remain relevant amidst competition with digital media. However, the analysis of listener satisfaction levels faces obstacles in the form of survey data imbalance, where certain categories have much less data than other categories. This causes the analysis results to be less representative, making it difficult for management to design strategies to improve broadcast quality. This study aims to overcome this problem by applying the Synthetic Minority Oversampling Technique (SMOTE) method to handle data imbalance, and the K-Nearest Neighbors (KNN) algorithm to classify listener satisfaction levels. SMOTE is used to generate synthetic data in the minority category, so that the survey data is balanced. Furthermore, KNN is used to classify satisfaction levels based on the balanced dataset. The results of the study show that the combination of SMOTE and KNN is able to provide more accurate and representative classification results. The information generated helps management in understanding the level of listener satisfaction as a whole, so that it can be used as a basis for designing strategies to improve the quality of broadcast programs