Query Performance Evaluation Over Health Data

Abstract

In recent years, there has been a significant increase in the number and variety of application scenarios studied under the e-health. Each application generates an immense data that is growing constantly. In this context, it becomes an important challenge to store and analyze the data efficiently and economically via conventional database management tools. The traditional relational database systems may sometimes not answer the requirements of the increased type, volume, velocity and dynamic structure of the new datasets. Effective healthcare data management and its transformation into information/knowledge are therefore challenging issues. So, organizations especially hospitals and medical centers that deal with immense data, either have to purchase new systems or re-tool what they already have. The new data models so-called NOSQL, its management tool Hadoop Distributed File Systems is replacing RDBMs especially in real-time healthcare data analytics processes. It becomes a real challenge to perform complex reporting in these applications as the size of the data grows exponentially. Along with that, there is customers demand complex analysis and reporting on those data. Compared to the traditional DBs, Hadoop Framework is designed to process a large volume of data. In this study, we examine the query performance of a traditional DBs and Big Data platforms on healthcare data. In this paper, we try to explore whether it is really necessary to invest on big data environment to run queries on the high volume data or this can also be done with the current relational database management systems and their supporting hardware infrastructure. We present our experience and a comprehensive performance evaluation of data management systems in the context of application performance.