Which Doctor to Trust: A Recommender System for Identifying the Right Doctors

Area of Science:

• Medical and Health Informatics
• Data Mining
• Artificial Intelligence

Background:

• Identifying key opinion leaders (KOLs) in medical and health informatics is critical for public health.
• Limited research exists on systematic methods for identifying disease-specific KOLs.

Purpose of the Study:

• To develop a novel recommender system for identifying KOLs for any specific disease using health care data mining.
• To leverage diverse professional data to pinpoint influential medical experts.

Main Methods:

• An unsupervised aggregation approach was used to integrate various ranking features.
• Professional footprints including publications, presentations, advocacy, and media exposure were collected for doctors.
• A recommender system was designed, implemented, and deployed.

Main Results:

• The system analyzed data from 2,381,750 doctors in China.
• Empirical results showed the system significantly outperformed benchmark systems.
• A real-world case study validated the method's applicability.

Conclusions:

• Doctor profiles and academic publications are essential for identifying KOLs in medical informatics.
• The deployed recommender system provides patients with authoritative doctor rankings for specific diseases.
• This facilitates better patient-doctor connections and access to expert medical advice.