Construction of Physician Group Profiles in Online Medical Consultation Platforms: A Tripartite Physician-Patient-Platform Perspective

Abstract

Constructing physician group profiles within online medical consultation platforms enables the precise characterization of multi-dimensional features of physician cohorts. This approach facilitates a better understanding of physicians’ behavioral patterns and service quality on these platforms, thereby providing support for service optimization. Drawing upon existing research on factors influencing patients’ physician selection behavior, a conceptual model for physician profiling was established from three dimensions: physician, patient, and platform. Using the concept lattice method, profiles for dermatology and venereology physicians on the Haodf.com platform were constructed. Based on the Hasse diagram, the physician cohort was categorized into four distinct types: high-value elite, multi-domain active, high-efficiency interaction, and potential-to-be-explored. Furthermore, implicit dependencies among attribute tags were uncovered through association rule mining. Adopting a tripartite “physician-patient-platform” perspective, we developed a group profile for physicians on online consultation platforms, which validates the efficacy of the concept lattice method for such profiling. The mined association rules further reveal the interrelationships within the “physician-patient-platform” data, offering data-driven support for optimizing physician management, enhancing service quality, and enabling precise recommendations on these platforms.