Design And Implmentation Of Medical Information System Based On Knowledge Graph

As a highly information-intensive and frequently updated field,medical information retrieval still stays in the Encyclopedia Search mode,medical practitioners retrieve information in a lot of lengthy professional documents.Meanwhile,false information and bad advertisements on the Internet also have a negative impact on the acquisition of medical knowledge by non-practitioners.Base on the above situations,a highquality medical knowledge service system needs to be established urgently.The existing medical knowledge graph has a large granularity and is mostly the medicine part of the general knowledge graph,which can not meet professional requirements.Therefore,this paper implements a fine-grained medical knowledge service system with medicine/disease as the core.The main research contents are as follows:1、This thesis studies the construction technology of knowledge graph in the medical field,and puts forward a complete construction scheme of knowledge graph.The structured,semi-structured and unstructured data on the Internet are processed via using web crawler technology and knowledge extraction based on deep learning.According to the authoritative classification methods of diseases and drugs,the medical ontology layer is designed,and the medical knowledge graph is constructed based on the processed data,which provides data support for the system functions,such as information retrieval,intelligent questionanswering and so on.2、To solve the problems of incomplete medical information,rapid update of new drugs and heavy memory burden of medical practitioners,a knowledge extraction algorithm for drug instructions is designed.A joint extraction model based on RoBerta-wwm+CLN is adopted to effectively handle the triple overlap.By comparing the performance of the proposed model and Bert+CLN joint extraction model with experiments,the effectiveness of the proposed model to improve the medical system performance is proved.3、Based on the above knowledge graph construction scheme and knowledge extraction algorithm,a medical knowledge service system is designed and implemented.Aiming at the problems of difficult information retrieval,serious fragmentation and uncertain correctness in the medical field,the system provides intelligent question-answering function based on question parsing,visualization of knowledge graph,encyclopedia retrieval and other functions.Finally,a comprehensive test of is carried out to prove the availability and stability of the system.Thus,this thesis proposes a construction scheme of knowledge graph in the medical field and a knowledge extraction algorithm for medical text based on deep learning technology.Finally,a medical knowledge service system supporting natural language question-answering is developed on the basis of knowledge graph.