A Modeling Approach to Big Data Based Recommendation Engine in Modern Health Care Environment

Over the past several years, data has been growing immensely in all business sectors. While many industries are successful in performing big data analysis to benefit from data sets, health care sector has started to take small steps to move forward. Health care providers and investors are actively investing in data analytical capabilities to successfully benefit from these data sets. This move will help them to have a better understanding of the complexity of changing health care environment. Our goals are to manage and integrate various unstructured health care big data sets in a secure environment, to generate useful knowledge from these unstructured data sets and to translate the knowledge into a working useful practical model. The main focus of this research project work is to build an application system for early identification of diseases. This application system can be a very helpful tool for the health care service providers to improve both overall quality and efficiency in the health care area. The application system is built, using Naïve Bayes(NB) classification algorithm running on top of Apache Mahout, to recommend the health conditions of users, readmission rates, treatment optimization, and adverse events. The existing health care approaches are mostly based on standard regression methods, which have limitations. Our focus research work will be on analyzing and using new big data methodologies. We use NB classification algorithm for diagnosing the diseases and providing necessary treatments suggestions. Once the disease is identified, delivering the correct care to the patients should optimize treatment cost. Also average life expectancy of people can be increased if they are treated with the appropriate care from early stages.