Healthcare Analytics with R and MongoDB using Social Media

Abstract

The recent advent of various social media platforms has opened new avenues of data collection and analysis. The unstructured social media data needs proper data classification for efficient healthcare analytics. R has latency and the latency is induced by need to 'load' 'offline' data files. This paper introduces a framework to create a model to offset the processing of streaming data in R. This paper demonstrates how the framework can work by analysing the frequency of the international classification of diseases (ICD-10) keywords as a part of healthcare analytics. The proposed framework offsets the work of resource intensive analytics tasks like dynamic querying, summation, aggregation, map-reduce by performing these at the NoSQL data store and R can use these pre-computed results to perform subsequent analytics. This paper also illustrates how efficiency can be achieved in processing streamed data in R by comparing processing times with and without use of the proposed model.