Data set operations to hide decision tree rules

Abstract

Abstract.1 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 cryptographic techniques -which restrict the usability of the data -since the raw data itself is readily available for public use. We show some key lemmas which are related to the hiding process and we also demonstrate the methodology with an example and an indicative experiment using a prototype hiding tool. INTRODUCTIONPrivacy preserving data mining [1] is a quite recent research area trying to alleviate the problems stemming from the use of data mining algorithms to the privacy of the data subjects recorded in the data and the information or knowledge hidden in these piles of data. Agrawal and Srinkant [2] were the first to consider the induction of decision trees from anonymized data, which had been adequately corrupted with noise to survive from privacy attacks. The generic strand of knowledge hiding research [3] has led to specific algorithms for hiding classification rules, like, for example, noise addition by a data swapping process [4]. A key target area concerns individual data privacy and aims to protect the individual integrity of database records to prevent the reidentification of individuals or characteristic groups of people from data inference attacks. Another key area is sensitive rule hiding, the subject of this paper, which deals with the protection of sensitive patterns that arise from the application of data mining techniques. Of course, all privacy preservation techniques strive to maintain data information quality.The main representative of statistical approaches [5] adopts a parsimonious downgrading technique to determine whether the loss of functionality associated with not downgrading the data, is worth the extra confidentiality. Reconstruction techniques involve the redesign of the public dataset [6][7] from the non-sensitive rules produced by algorithms like C4.5 [8] and RIPPER [9]. Perturbation based techniques involve the modification of transactions to support only non-sensitive rules [10], the removal of tuples associated with sensitive rules [11], the suppression of certain attribute values [12] and the redistribution of tuples supporting sensitive patterns so as to maintain the ordering of the rules [13].In this paper, we propose a series of techniques to efficiently protect the disclosure of sensitive knowledge patterns in classification rule mining. We aim to hide sensitive rules without 1 School of Science and Technology, Hellenic Open University, Patras, Greece, email: kalles@eap.gr, verykios@eap.gr, georgios.feretzakis@ac.eap.gr 2 Epignosis Ltd, Athens, Greece, email: papagel@efrontlearning.net compromising the information value of the entire dataset. After an expert selects the sensitive rules, we modify class labels at the tree node corresponding to the tail of the sen...

On Using Linear Diophantine Equations for in-Parallel Hiding of Decision Tree Rules

“…These two passes help us firstly hide all sensitive rules and secondly keep the sanitized tree close to the structure of the original decision tree. We note that the above two procedures had been fully described in previously published works [16,17].…”

“…In our previous published work [16,17], the above procedure was greedy, mostly solving the issue for only one (tree) level of nodes, which often resulted in a non-minimum number of added instances, whereas a look ahead based solution would be able to take into account all levels up to the We used the information gain as the splitting heuristic. In order to hide the leaf which corresponds to the nine positive instances (to the right of N0), we changed the nine positive instances to negative ones and denoted this operation by (−9p,+9n).…”

“…In our previous published work [16,17], the above procedure was greedy, mostly solving the issue for only one (tree) level of nodes, which often resulted in a non-minimum number of added instances, whereas a look ahead based solution would be able to take into account all levels up to the…”

“…In our previous published work [16,17], the above procedure was greedy, mostly solving the issue for only one (tree) level of nodes, which often resulted in a non-minimum number of added instances, whereas a look ahead based solution would be able to take into account all levels up to the root. Furthermore, the new ratios of the nodes (p:n) were not the same as they were before the change, thus propagating ratio changes, whose impact could only be quantified in a compound fashion by inspecting the final tree, hampering our ability to investigate the behaviour of this heuristic in a detailed fashion.…”

“…In our previously published works [16,17], we proposed a series of techniques to adequately protect the disclosure of sensitive patterns of knowledge in the mining of classification rules. We aimed to hide sensitive rules without compromising the information value of the whole dataset.…”