<![CDATA[Determining research topics for final projects that align with students' competencies and interests remains a challenge in academic management because the process is often influenced by subjective factors. This situation can lead to various impacts, such as inappropriate research topics, delays in study completion, and imbalances in the guidance workload among supervisors. This study aims to cluster the research interests of students preparing their final projects in the Computer Science Study Program at the State Islamic University of North Sumatra (UINSU) using the Fuzzy C-Means (FCM) algorithm. The data used were obtained from a student interest questionnaire from the 2020–2022 intake and historical data on final project titles. Furthermore, the text data underwent preprocessing and was converted into numerical form using the Term Frequency–Inverse Document Frequency (TF-IDF) method. The FCM algorithm was then used to form research interest clusters with fuzzy membership degrees. Based on the results of the cluster quality evaluation, it was found that the most optimal number of clusters was six, with a Silhouette Index value of 0.6311 and a Davies–Bouldin Index of 0.5505, which indicates that the cluster structure formed is classified as good. The clustering results indicated that student interests were dominated by Software Engineering and Artificial Intelligence, with a fairly high degree of overlap. This study combines student interest questionnaire data and historical final project title data, represented using TF-IDF and clustered using the Fuzzy C-Means algorithm to map multidimensional research interests. The results suggest that this approach provides a more objective basis for identifying students’ research tendencies and can support topic recommendation systems and academic supervision planning.]]>
Copyrights © 2025
