Providing k-anonymity in data mining

Friedman, Arik; Wolff, Ran; Schuster, Assaf

doi:10.1007/s00778-006-0039-5

Cited by 104 publications

(61 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…These methods transform the dataset into k indistinguishable records from each other. Mostly, k-anonymity concept is employed by the PPDM algorithms so as to assure privacy [5]. But this method is not suitable for high dimensional data and to search out optimal k-anonymous datasets over generalization and is esteemed as NP-Hard [6] [7].…”

Section: Literature Reviewmentioning

confidence: 99%

Untitled

2016

IJSC

View full text Add to dashboard Cite

Privacy preservation is that the most targeted issue in information publication, because the sensitive data shouldn't be leaked. For this sake, several privacy preservation data mining algorithms are proposed. In this work, feature selection using evolutionary algorithm and data masking coupled with slicing is treated as a multiple objective optimisation to preserve privacy. To start with, Genetic Algorithm (GA) is carried out over the datasets to perceive the sensitive attributes and prioritise the attributes for treatment as per their determined sensitive level. In the next phase, to distort the data, noise is added to the higher level sensitive value using Hybrid Data Transformation (HDT) method. In the following phase slicing algorithm groups the correlated attributes organized and by this means reduces the dimensionality by retaining the Advanced Clustering Algorithm (ACA). With the aim of getting the optimal dimensions of buckets, tuple segregating is accomplished by Metaheuristic Firefly Algorithm (MFA). The investigational consequences imply that the anticipated technique can reserve confidentiality and therefore the information utility is additionally high.Slicing algorithm allows the protection of association and usefulness in which effects in decreasing the information dimensionality and information loss. Performance analysis is created over OCC 7 and OCC 15 and our optimization method proves its effectiveness over two totally different datasets by showing 92.98% and 96.92% respectively.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Untitled

2016

IJSC

View full text Add to dashboard Cite

show abstract

“…One of the advantages of generalization over other disclosure limitation techniques is that it preserves the truthfulness of the information. However, a major drawback of existing generalization techniques is that domain hierarchy trees are required for every quasiidentifier attribute on which k-anonymity is to be applied [4], [5], [6], [7], [8], [9], [10]. These domain hierarchy trees are generated manually by the user before applying the generalization process.…”

Section: Related Workmentioning

confidence: 99%

“…This section presents the results of various issues that were examined in regard to the proposed algorithm: (1) verification of the proposed hybrid approach for achieving k-anonymity without reasonable loss of classification accuracy; (2) comparison of kACTUS-2 with TDS, TDR and kADET algorithms presented in [6], [9] and [10] in terms of classification accuracy, information loss and statistical significance; (3) comparison of kACTUS-2 with former algorithm kACTUS in terms of classification accuracy, information loss, statistical significance and data loss.…”

Section: Experimental Evaluationmentioning

confidence: 99%

kACTUS 2: Privacy Preserving in Classification Tasks Using k-Anonymity

Kisilevich

Elovici²,

Shapira³

et al. 2009

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. k -anonymity is the method used for masking sensitive data which successfully solves the problem of re-linking of data with an external source and makes it difficult to re-identify the individual. Thus kanonymity works on a set of quasi-identifiers (public sensitive attributes), whose possible availability and linking is anticipated from external dataset, and demands that the released dataset will contain at least k records for every possible quasi-identifier value. Another aspect of k is its capability of maintaining the truthfulness of the released data (unlike other existing methods). This is achieved by generalization, a primary technique in k -anonymity. Generalization consists of generalizing attribute values and substituting them with semantically consistent but less precise values. When the substituted value doesn't preserve semantic validity the technique is called suppression which is a private case of generalization. We present a hybrid approach called compensation which is based on suppression and swapping for achieving privacy. Since swapping decreases the truthfulness of attribute values there is a tradeoff between level of swapping (information truthfulness) and suppression (information loss) incorporated in our algorithm.We use k -anonymity to explore the issue of anonymity preservation. Since we do not use generalization, we do not need a priori knowledge of attribute semantics. We investigate data anonymization in the context of classification and use tree properties to satisfy k-anonymization. Our work improves previous approaches by increasing classification accuracy.

show abstract

“…In the proposed K-anonymity method, CBK(L,K)-anonymity algorithm is used to solve the problem of privacy information leakage in data publishing.The main idea is anonymizing the data set by K-clustering based on influence matrix of background knowledge, it can make anonymous data effectively resist background knowledge attack and homogeneity attack, and can solve diversity of sensitive attribute. [6]. …”

Section: B K-anonymity Clustering Methods For Effective Data Privacy mentioning

confidence: 99%