Automated Faculty Evaluation and Ranking System: Utilizing OCR, NER, and Decision Tree for a Web-Based Evaluation System

Abstract

The Pamantasan ng Lungsod ng Maynila (PLM) a first and only chartered and autonomous university funded by a city government of Manila, faces challenges in faculty evaluation due to manual processes. Initially, the manual computation for faculty classification, ranking, and promotion result in errors and extended processing times. The manual faculty selection process is time-consuming and susceptible to human errors coupled with the absence of an automated decision support system makes it difficult for evaluators to streamline the faculty evaluation process. EduRate is a web-based application that utilizes machine learning, including Optical Character Recognition, Named Entity Recognition, and Decision Tree that automates the evaluation of faculty applicants, streamlining processes, improving workflow efficiency, and reducing administrative burdens. The researchers analyzed the data gathered using mean, utilizing the ISO/IEC 25010 standard. The participants were full-time and part-time faculty members from the College of Engineering (CoE) at PLM, selected using purposive sampling. Subsequently, applying Slovin's formula with a margin of error of 5% which was 97 respondents. Based on the summary of findings from a 5-point Likert scale, the research found that functional suitability (mean of 1.87), usability (mean of 1.60), reliability (mean of 1.86), security (mean of 1.81), and maintainability (mean of 1.89) were all rated as extremely to very effective. The research findings highlight the potential of EduRate to significantly enhance the quality of faculty evaluation and decision-making in hiring and promotion. Moreover, this can help institutions save resources, efficiently on evaluators’ time, and transparent reports.