<![CDATA[The objective of this research is to design and implement a performance evaluation system that combines Machine Learning for data processing, predictive modeling, and pattern recognition with the GAP method to measure discrepancies between expected competencies and actual performance. Eight primary criteria were cooperation, communication, initiative, alertness, discipline, leadership, problem solving, time usage each consisting of several sub-criteria. The study involved 18 lecturers, and the evaluation was conducted using a web-based decision support system equipped with machine learning models trained to classify performance levels and identify underlying patterns within the assessment data. System usability was examined through four categories: ease of use, completeness, accuracy, and interface composition. The results show that the integrated system successfully identified the highest-performing lecturer (Lecturer 7) with a score of 6.1801, followed by Lecturer 12 with 4.9314 and Lecturer 4 with 4.1157. Usability testing also yielded positive outcomes, with scores of 89% for ease of use, 87% for completeness, 90% for accuracy enhanced through machine learning validation and 88% for interface composition. These results produced an overall average of 88%, classifying the system as Very Worthy. In conclusion, integrating Machine Learning and GAP Analysis in a web-based DSS significantly improves the effectiveness and efficiency of lecturer performance evaluation. The system accelerates data processing, enhances assessment quality, and strengthens decision-making through predictive analytics and automated classification. This framework offers a valuable reference for future performance evaluations in higher education institutions seeking accountability, transparency, and data-driven decision-making.]]>
