On Using Linear Diophantine Equations for Efficient Hiding of Decision Tree Rules
Georgios Feretzakis, Dimitris Kalles, Vassilios S. Verykios · 2018
Data sharing among organizations has become an increasingly common procedure in several areas like advertising, marketing, e-commerce and banking, but any organization will probably attempt to keep some patterns as hidden as possible when it shares its datasets with others. This paper focuses on preserving the privacy of sensitive patterns when inducing decision trees. We adopt a record augmentation approach for hiding sensitive classification rules in binary datasets. Such a hiding methodology is preferred over other heuristic solutions like output perturbation or crypto-graphic techniques - which restrict the usability of the data - since the raw data itself is readily available for public use. We propose a look-ahead approach using linear Diophantine equations in order to add the appropriate number of instances while maintaining the initial entropy of the nodes. This technique can be used to hide one or more decision tree rules in an optimal way.