Data clustering based anonymization approaches for big data privacy

Abstract

In recent years, with the emergence and usage of new systems and Internet technologies, people get connected with each other through various cyber society components. This interaction of people led to the accumulation of huge amount of data generated from different sources including social data, machine data, and transactional data. Generally, the size of the data is ranging from a few dozen terabytes to many zettabytes of data indicated by International Data Corporation. newlineBig data specifically refers to data sets that are so large in size as well as complex that inundates a business on a daily basis. It is a data asset with a lot of volume, speed, and variety for providing valuable insight and decision making through cost-effective and innovative data processing. newlineBig data analytics is helpful in various industries like medical fields, banking sectors, network security, and social media to extract meaningful information for making better decisions about future. To examine big data, a variety of software tools, as well as advanced analytics disciplines such as predictive analytics, text analytics, and statistical analysis are used. newlineThe Electronic Health Record (EHR) maintained at the hospitals have many useful resources for the prevention of disease, health information exchange, and for making useful medical decision. Data owners publish or outsource this information for better profits. However, EHR data contain sensitive information about the patients used for medical diagnosis and medication. According to Health Insurance Portability and Accountability Act (HIPAA 1999) privacy law, the medical information kept at any medical health center should be kept confidential. newline newline