Penerapan Data Mining Untuk Clustering Data Pencari Kerja dengan Menggunakan Algoritma K-Means

Abstract

Job seekers applying for employment must adhere to the policies of the desired company. A current and relevant issue is the high number of job seekers who struggle to secure employment due to various factors, including educational levels and job availability in different regions. Job applicants across various companies often possess diverse age and educational qualifications, particularly in the Bantul Regency, D.I Yogyakarta. The data utilized in this research is sourced from the Bantul District Manpower and Transmigration Office. Data processing is carried out using Data Mining or Knowledge Discovery in Database, employing the clustering method with the K-Means algorithm to categorize job seeker data into desired clusters. The aim of this study is to group job seeker data in the Bantul Regency and segregate them into groups with distinct characteristics. Testing is conducted using the RapidMiner software, a Data Mining tool with various available methods. The grouping of job seeker data is expected to assist the government in evenly distributing jobs to job seekers and reducing the accumulation of job seeker data annually. Additionally, it can aid the Bantul District Manpower and Transmigration Office in planning and allocating resources according to needs. The implementation results of the K-Means algorithm on job seeker data, tested from k=2 to k=10, yield the best cluster at k=2 with a Davies Bouldin Index (DBI) value of 0.523. Cluster_0 consists of 214 members, while Cluster_1 consists of 36 members. Therefore, the smaller the Davies Bouldin Index (DBI) value, the higher the similarity level of data within one cluster.

Keywords: Data Mining, Job Seekers, RapidMiner, Clustering, K-Means Algorithm