<![CDATA[Manual assessment of practicum report margins often faces challenges due to variations in document layout and the large number of reports that must be reviewed, resulting in an inefficient and subjective evaluation process. Therefore, this study aims to develop a prototype that serves as a guideline for automated practicum report assessment. The proposed research introduces a hybrid model integrating Fuzzy Logic and Naive Bayes. Evaluation results show that the model without fuzzy achieved an accuracy of 73.33% but exhibited bias toward the majority class, with low recall for the "Rejected" class. In contrast, the Fuzzy Naive Bayes model improved accuracy to 80% and produced more balanced classification performance, with significant increases in recall and F1 Score for the minority class. The integration of fuzzy logic effectively enhances the detection of margin non-compliance.]]>
Copyrights © 2025
