A machine learning-based hybrid recommender framework for smart medical systems.

Area of Science:

• Medical Informatics
• Artificial Intelligence
• Big Data Analytics

Background:

• Traditional medical systems face challenges in accurate service evaluation and personalized doctor recommendations.
• Existing recommender systems often lack the sophistication to handle complex medical data and patient needs.

Purpose of the Study:

• To develop a hybrid recommender framework for smart medical systems to improve service level evaluations and doctor recommendations.
• To enhance the accuracy and efficiency of medical appointment platforms through advanced AI techniques.

Main Methods:

• A registration review system using big data and deep learning for improved institutional evaluations.
• A doctor recommendation model employing term frequency-inverse document frequency (TF-IDF), modified cosine similarity, K-means clustering, alternating least squares (ALS) matrix decomposition, and user collaborative filtering.
• Construction of a patient symptom vector space for personalized score calculation.

Main Results:

• The proposed registration review system demonstrated higher accuracy than conventional evaluation methods.
• Significant improvements in precision and recall rates were observed in doctor recommendations compared to existing approaches.
• The hybrid framework provides a practical solution for department triage and doctor recommendation.

Conclusions:

• The hybrid recommender framework offers a robust and accurate solution for smart medical systems.
• The integration of big data, deep learning, and collaborative filtering enhances patient experience and optimizes medical resource allocation.
• This approach represents a significant advancement in medical appointment platform functionality.