A recommendation model based on profile enhancement for personalized medical expert in online health communities

Inefficient matching between medical experts and health advice seekers remains a challenge in online health communities. This study constructs the Recommendation Model with Profile Enhancement (RMPE), which integrates service quality and user sentiment dimensions to enhance service exchange efficiency, promote value co-creation, and provide algorithmic support and managerial insights for the ecological governance of online health communities.

The RMPE workflow consists of three stages: extracting dynamic expert profiles via topic modeling, quantifying users' subjective experiences through sentiment analysis, and completing personalized recommendations via collaborative filtering. The model computationally integrates Service-Dominant Logic and Information Ecology Theory into the recommendation process.

Empirical analysis based on data from Haodf.com demonstrates that RMPE achieves a recommendation accuracy of 88.1%, significantly outperforming traditional baseline models, and INTERNAL/BUSINESS successfully identifies over 80% of users' final choices within the Top-30 recommendation list.

The findings validate the technical effectiveness of the profile enhancement pathway and provide a framework for quantitatively analysing value co-creation processes in online health communities. The proposed approach provides both Theoretical and practical value for platform governance by supporting resource allocation optimization, community culture cultivation, and the sustainable development of the health community ecosystem.