Personalized Doctor Recommendation Method Based on Patient Online Consultation Text

Provide personalized doctor recommendation for patients on online medical platform, and improve the accuracy of doctor recommendation. Using Word2vec algorithm, TF-IDF algorithm, and cosine similarity theorem, form a candidate doctor recommendation set based on similar patients and similar doctors, and the two recommendation methods are combined to form a doctor recommendation set based on mixed recommendation. Taking the data on www.haodf.com as an example, the consultation texts for different patients matched suitable doctor resources for them, and the results of mixed recommendation are relatively accurate and have certain practical significance.