<![CDATA[This research focuses on mapping students' final assignment topics in the Computer Science Study Program at the North Sumatra State Islamic University (UINSU) using the Latent Dirichlet Allocation (LDA) algorithm. The background to this research stems from the need to understand research developments and trends in the collection of submitted final assignments, which can provide an overview of academic trends and developing research areas. However, manual clustering of these topics is often a challenge due to the large data volume and complexity of the content. The Latent Dirichlet Allocation (LDA) algorithm offers a solution with its ability to automatically identify hidden topic structures in text documents. The aim of this research is to reveal dominant themes and topic patterns that appear in students' final assignments, so as to provide deeper insight into the research focus area. The research methodology includes collecting data from various final projects, preprocessing the data to reduce noise and redundancy, and applying the LDA algorithm for topic extraction. The research results show that the LDA algorithm is effective in mapping the topics of students' final assignment titles at UINSU. By using 1000 iterations of the LDA process on 774 final assignment titles, it was found that the most optimal topic division was 7 topics with a coherence score of 0.4011. These topics are visualized through word clouds and word lists, which facilitate understanding and thematic interpretation. It is hoped that these conclusions will provide useful insights into student research trends, facilitate assessment of the quality and relevance of topics, and support the development of better academic curricula in higher education institutions.]]>
Copyrights © 2024
