Privacy Preserving Data Mining in Electronic Health Record using K- anonymity and Decision Tree

In this paper, we present an accurate and efficient privacy preserving data mining technique in Electronic Health Record (EHR) by using k –anonymity and decision tree C4.5 that is useful to generate pattern for medical research or any clinical trials. It is analyzed that anonymization offers better privacy rather than other privacy preserving method like that randomization, cryptography, perturbation and encryption methods. KAnonymity is useful for prevention of identity of patient in medical research; it is useful to avoid linking attack using suppression and generalization process. The objective of this research work is to propose and implement privacy preserving data mining approach, which best suits with EHR systems without impeding the flow of control. The data stored in EHR system is highly confidential which contains information about patient disease. To ensure confidentiality, we are using anonymization of identity revealing attribute before publishing it for other utility purpose. The experimental results show the validity of the proposed approach. Our approach is useful to preserve utility and privacy in healthcare datasets.