Research on the Application of Semantic Network in Disease Diagnosis Prompts Based on Medical Corpus

Abstract

Portion of the causes of medical errors in outpatient clinics are incorrect treatment resulting from misdiagnosis. Misdiagnosis between diseases is often caused by similar and indistinguishable symptoms. Currently, disease knowledge and related symptom words that are prone to misdiagnosis are scattered in various medical literature or open online databases. Therefore, it is possible to merge symptom words of related diseases, build an ontology based on the semantic relationship between symptoms, and associate the association between diseases and symptoms. This project finally established the "Disease Symptoms" Semantic Network (DSSN). Using DSSN as a basic data set can serve as prompts for diseases that are easily misdiagnosed, assisting doctors in accurately diagnosing diseases. This plays a significant role in reducing the misdiagnosis rate.