Heuristic Based Approach for Privacy Preserving in Data Mining
Patel Shreya, Aniket Patel · Zenodo (CERN European Organization for Nuclear Research) · 2018
Data Mining has been the foremost researched space for researchers because of the possibilities of extension at every application of it. Once the information becomes massive in volume, several issues strike for security and privacy breach. If the data changes, it'd be mandatory to rescan the database that results to long computation time and inability to promptly reply to the user. Some applications like sharing of such knowledge to a selected user that have threats of preserving the original data so that the injection of such data can be prohibited. Thus it's a timely need to secure the information whereas handling them to the known or unknown users. Such troubles prompted the advancement of Privacy Preserving Data Mining (PPDM) Techniques. Primary objective is to accomplish harmony between privacy preservation and knowledge discovery and hiding the data from attacker. Privacy Preserving has become a crucial issue within the development progress of Data Mining techniques. Methods like k-anonymity along with the hybrid approach of l-diversity and t-closeness. Experimental outcomes shows that the approach not solely preserve's data privacy however one will get better accuracy with minimum loss of data.