Improved method of selecting data in a nonrelational database

Abstract

People surround themselves with data in many ways, which evokes the need for correct ways of storing the data. Nowadays, the trend tends to lean in favor of data storing in nonrelational (or NoSQL) databases. These databases are used in various user applications, which need a huge volume of the data highly accessible and do not require big data consistency. The problem of the data growth and its storing in the nonrelational databases results in the decreasing efficiency of searching in data. In the paper, we present use of very popular in-memory database in order to help us with this lack of efficiency of the data searching. This paper examines the data searching in applications hosted by Amazon cloud service while using nonrelational database DynamoDB. We develop new procedures to provide faster response to user and to obtain the data using nonrelational database DynamoDB. These procedures provide the queried data and subsequently, transfer them into the memory. The given procedures are based on two methods. The first method is a recognition of values, the user refers to and the provision of this data to the in-memory database. The second method is related to the automatic storing of the data transferred to the in-memory database. We perform number of experiments, which are describing a limitation of efficiency/inefficiency from a perspective computational time.