The Association Rule Methods and K-Means Clustering For Optimization Mapping Of New Students Admission

Educational service providers deliver information about student admission by installing banners, distributing brochures, and providing information directly to prospective new students annually. The target result can be outside expectations because there is no data analysis related to optimizing the support of the system. This study proposes the K-Means algorithm optimized with the Association Rules method to map the interests of prospective students in selecting the study program. The data collection method uses an explanatory research survey and correlational study with primary data from the questionnaires and documents for new students in the year 2022. The accuracy results of this study using the K-Means algorithm process to calculate the silhouette score, elbow methods, cross-combination experiments, and data sharing are 95.00%. The optimization accuracy with mapping relationships between variables using frequent itemset and Association Rules increased to 100%. The most significant factors influencing the student selection of study programs are religion 100%, transportation 85.71%, gender 83.33%, school status 80.95%, regency 73.81%, school origin 71.43%, the district of the school close to student life 66.67%, residence located at the parent's home 64.29%, place of birth 61.90%, and job future ambition 61.90%.