Design of the Korean medicine symptom diagnosis system using Word2Vec

Abstract

This paper presents the design of the Korean medicine symptom diagnosis system. In the ordinary korean medicine symptom diagnosis situation, the patient describes their symptoms and the korean medicine doctor makes a diagnosis based on the symptoms. In order to develop the symptom diagnosis system that can diagnose itself, the system has to understand human natural language. In this paper, we use morpheme and synonyms to give understandable information to the system in terms of the information retrieval system. We design database schema that consists of NoSQL document-oriented databases-MongoDB to get a better performance at unstructured and semi-structured data. We also use Word2Vec that is kind of word embedding to solve synonyms problem.