ANALYSIS OF THE STORAGE PLATFORM FOR CREATING AN INFORMATION REPOSITORY OF MEDICAL RESEARCH

Анотація

This article discusses the analysis of types of non-relational databases, namely these document-oriented databases, "key values" databases, columnar databases, graphic databases for storing medical research. To ensure an understanding of the structure of medical data storage, an information model for the storage of medical records was created. After a detailed analysis of the advantages and disadvantages of non- relational database types, this paper identifies the features of the document-oriented database type and created this model for further analysis of general optical information analysis for medical research. Due to the choice of a document-oriented type of database, a platform for storing MongoDB medical data based on a document-oriented database was chosen. MongoDB is a cross-platform open source document management (DBMS) system. The choice of the document-oriented DBMS is based on the fact that, unlike relational analogues, there are no restrictions on the set of fields in the documents. MongoDB meets all the requirements described in these article, as well as these documents - focused DBMS. MongoDB was chosen precisely because of the popularity, similarity of the approach to data storage with relational databases and BSON data storage format. To summarize the entire study, it should be noted the following research results: 1. Created and described an information model for the storage of medical research. 2. On the basis of the qualitative analysis of four types of non-relational models of databases, designed by a documentary-oriented model of databases to create an information analysis of medical research. 3. Based on the selected type of databases, it is proposed to choose the document-oriented database management system - MongoDB Big Data Platform for medical data storage.