Pertinent Exploration of Privacy Preserving Perturbation Methods

Abstract

Digital era generates a huge amount of data in many sectors like education, medical, banking, business, marketing, etc. which can be used for research motive, analysis, prediction of trends, statistics, etc. Data mining techniques are useful in finding patterns, trends, and knowledge from such huge data. The data holders are not ready to share data because there are chances of privacy leakage. Sharing of such data immensely helps researchers to obtain knowledge from it, especially medical data. Privacy preserving data mining is one way where researchers will get mine data for gaining knowledge without breaching the privacy. In the medical sector there is a branch called the mental health section, where high confidentiality of data is maintained and is needed. Owners are not ready to share data for research motives. Mental health is nowadays a topic that is most frequently discussed when it comes to research. PPDM allows sharing data with the researcher, where the privacy of data is maintained by using perturbation techniques giving relief to doctors (owner of data). The current paper experiments and analyses different perturbation methods to preserve privacy in data mining