Master Data Monitoring And Analysis System For Large-scale Medical Institutions

With the development of science and technology and the progress of medical and health conditions,the number of pharmacies,hospitals and clinics in major cities in China has grown rapidly,and the pressure of master data identification caused by the sharp increase in the number of "aliases" received by data departments in the medical industry has become increasingly prominent.In order for data cleaners to effectively distinguish and accurately judge the corresponding relationship between terminal information and master data,the similarity recognition algorithm and information monitoring system for text recognition have become one of the research focuses.As the representative master data,the "alias" data of the medical industry has gradually formed large-scale basic data with the continuous expansion of the data volume with the accumulation of business.Therefore,relevant enterprises urgently need an efficient master data management scheme to deal with the current difficulties.In this paper,based on the accumulated master data related to the medical industry,a fast search engine is established in three aspects: rapid retrieval and similarity recognition of large-scale master data,conflict monitoring and control of master data management,and analysis and display of result data,and the similarity recognition algorithm of master data applicable to the medical industry is derived.At the same time,the design of monitoring and analysis system in the process of master data management is completed.The main work contents of this paper are as follows: 1)Due to the huge amount of master data related to the medical industry in the production environment and usually containing rich medical knowledge,this paper first preprocesses the master data related to the medical industry,and then builds a set of data structures and storage methods for the characteristics of the master data of the medical industry.2)For the query problem of large-scale master data,we first consider the fast retrieval method of word segmentation index;Secondly,in the selection of master data fuzzy recognition,this paper proposes a text similarity recognition algorithm based on semantic classification,which combines Ro BERTa model and text difference set conflict detection.3)In order to solve the problem that text similarity recognition cannot be automatically recognized due to insufficient similarity,this paper proposes to use external combined query as a supplement to the recognition algorithm,and at the same time,gives consideration to the conflict between combined query and existing data,and proposes a conflict monitoring and resolution mechanism.4)A set of master data management system for medical institutions in the field of life science has been built,which mainly includes data storage,data retrieval,similarity scoring,analysis and display and other functions,and can provide data security for managers.To sum up,this paper has completed the establishment of master data storage mode for large-scale medical institutions,and applied the fuzzy recognition algorithm and conflict resolution mechanism for master data of medical institutions.The system has achieved the expected effect in the test and experiment of more than 1.5 million real medical institution master data of enterprises,and has practical significance.