Application of Text Mining and K-Medoids for Job Clustering Based on Job Description Analysis

The massive growth of job description data poses challenges in manual job classification due to the unstructured nature of the data. This study aims to develop a job cluster grouping model using a text mining approach and the K-Medoids algorithm. Job description data was obtained from the O*NET Online website and supplemented with data from various prominent job portals in Indonesia. The data was then processed through the stages of tokenization, stopword removal, lemmatization, and vector representation using SentenceTransformer. The clustering process was performed using the K-Medoids algorithm with the Euclidean distance metric, and evaluation was conducted using the Silhouette Score and Davies-Bouldin Index. The results showed that the model produced a sufficiently representative classification with a Silhouette Score of 0.48770 and a Davies-Bouldin Index of 0.815, demonstrating better performance compared to the Agglomerative Clustering method. This approach is effective for supporting automated and data-driven human resource management.