Enterprise AI Analysis

Project GEDI: Promoting Gender Equity and Diversity in Classroom Settings using Artificial Intelligence

Pilita A Amahan, Information Technology, Occidental Mindoro State College

Nimfa B Pastrana, Graduate School, Occidental Mindoro State College

Artemio M Gonzales Jr, Midwifery Department, Occidental Mindoro State College

This study investigates gender-based disparities in classroom participation, feedback sentiment, and instructional content using ethically collected data from 300 students. Three equity indicators: underrepresentation, stereotyped feedback, and material bias were flagged through a structured, rule-based annotation framework. Analysis revealed that female and nonbinary students consistently experienced reduced engagement, neutral feedback, and exposure to biased materials, while male students received higher participation scores and affirming feedback without content bias. To enhance interpretative depth, findings were further analyzed using Transformer-based sentiment models (BERT and FAIRBert), which were benchmarked against the rule-based guidelines. Composite indices such as the Gender Responsiveness Score underscored persistent structural inequities across identity groups. These diagnostic outcomes directly informed the design and development of Project GEDI, an AI-powered prototype that translates classroom data into real-time bias evaluations and pedagogical recommendations. More than an evaluation tool, GEDI functions as a dynamic equity monitor that supports inclusive teaching practices. This research demonstrates the effectiveness of combining semantic annotation, AI modeling, and human intervention to operationalize inclusive education. Future work may expand GEDI's scope to support longitudinal tracking, subject-specific calibration, and adaptive interventions across diverse learning environments.