Abstract

The rise of digital health platforms has reshaped the way individuals manage chronic health conditions, providing tools for peer support and personalized care. However, these platforms face significant challenges in content moderation and peer-matching systems, critical for fostering trust and meaningful engagement. Med-Mingle addresses these gaps by integrating advanced natural language processing models, such as BERT, for real-time, context-sensitive moderation, and hybrid recommendation algorithms for dynamic, user-centered peer matching.

The platform leverages collaborative filtering and demographic-based systems enhanced with machine learning to create impactful peer connections while prioritizing user safety and inclusivity. Ethical considerations, including data privacy, fairness, and transparency, are central to its design, ensuring compliance with regulatory standards. By combining robust AI models with a user-focused approach, Med-Mingle demonstrates the potential of digital platforms to foster supportive and safe health communities.

This study offers a multi-phase development framework, encompassing data collection, fine-tuning transformer models, and implementing scalable hybrid systems. The findings underscore the transformative role of AI in health engagement and provide a foundation for future innovations in digital health platforms, enabling safer, more inclusive, and engaging user experiences.

Advisor

Guarnera, Heather

Department

Computer Science

Publication Date

2025

Degree Granted

Bachelor of Arts

Document Type

Senior Independent Study Thesis

Download

Share

COinS
 

© Copyright 2025 Linda Berhe