Optimization-based k-anonymity algorithms

Liang, Yuting; Samavi, Reza

doi:10.1016/j.cose.2020.101753

Cited by 24 publications

(20 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The anonymized database could be studied in place of the original database. Some common data anonymization models to prevent privacy disclosure include k-anonymity [10][11][12][13][14], l-diversity [9], t-closeness [15] and -presence [16]. a) k-anonymity: k-anonymity was developed to address identity disclosure.…”

Section: Data Anonymizationmentioning

confidence: 99%

“…In k-anonymity, any individual cannot be reidentified from the published data with a probability of higher than 1/k. Other variations of k-anonymity include clustering anonymity [11], distribution-preserving k-anonymity [12], optimizationbased k-anonymity [13], -sensitive-k-anonymity [14], (X,Y)-anonymity [17], (α, k)anonymity [18], LKC-privacy [19] and random k-anonymous [20] which prevent identity disclosure by hiding the record of a target in an equivalence class of records with the same QID values. Although k-anonymity model protects against identity disclosure, it is vulnerable against attribute disclosure.…”

Section: Data Anonymizationmentioning

confidence: 99%

See 1 more Smart Citation

Privacy-preserving healthcare informatics: a review

Chong

2021

ITM Web Conf.

View full text Add to dashboard Cite

Electronic Health Record (EHR) is the key to an efficient healthcare service delivery system. The publication of healthcare data is highly beneficial to healthcare industries and government institutions to support a variety of medical and census research. However, healthcare data contains sensitive information of patients and the publication of such data could lead to unintended privacy disclosures. In this paper, we present a comprehensive survey of the state-of-the-art privacy-enhancing methods that ensure a secure healthcare data sharing environment. We focus on the recently proposed schemes based on data anonymization and differential privacy approaches in the protection of healthcare data privacy. We highlight the strengths and limitations of the two approaches and discussed some promising future research directions in this area.

show abstract

Section: Data Anonymizationmentioning

confidence: 99%

Section: Data Anonymizationmentioning

confidence: 99%

Privacy-preserving healthcare informatics: a review

Chong

2021

ITM Web Conf.

View full text Add to dashboard Cite

show abstract

“…2. This error, defined for a couple (o, d), corresponds to the individual information loss [2], which independently penalize generalization of each attribute. For OD trips, we only have two attributes namely their origin and destination and we may measure by |o| and |d| their spatial generalizations in number of tiles.…”

Section: Problem Settingmentioning

confidence: 99%

“…Data protection approaches rely mostly on k-anonymization, which is achieved when all users in the data are indistinguishable from at least k − 1 other users. k-anonymization is usually attained through generalization and suppression [1,2], i.e., replacing values with less specific but semantically consistent values shared with other users, and deleting outlier users, respectively. However, k-anonymization of whole trajectories proves to be difficult to achieve for k > 5 and struggle to offer a truly foolproof anonymization [3].…”

Section: Introductionmentioning

confidence: 99%

“…The aggregation of origins gives a first spatial partition for origin areas, each origin being associated to a second spatial partition for destination areas. For each origin, finding the spatial partition for destinations that minimizes a generalisation error [2] reduces the anonymization problem to a tree-knapsack optimisation problem (TKP), or ordered knapsack problem [10]. Resulting volumes v od from an origin o to a destination d are considered anonymized if v od > k and are suppressed if v od ≤ k. These simplifications let us find the least destructive anonymization as measured by the generalisation error for high values of k. The rest of this paper is organized as follows: in Section 2, we formalize the problem and propose an algorithm to solve it.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Lightweight Approach for Origin-Destination Matrix Anonymization

Matet¹,

Côme²,

Furno³

et al. 2021

ESANN 2021 Proceedings

View full text Add to dashboard Cite

Personal trajectory data are becoming more and more accessible and have a high value in transport planning and mobility characterisation, at the cost of a risk for user's privacy. Addressing this risk is usually computationally expensive and can lead to losing most of the data utility. We explore a new, light-weight approach to Origin/Destinationmatrix anonymization that is easily scalable. We apply it to trip records from New York City Taxi and Limousine Commission (TLC) to illustrate how it can combine foolproof anonymity with a good spatial precision for a reasonable computational cost.

show abstract

A clustering‐based anonymization approach for privacy‐preserving in the healthcare cloud

Abbasi

Mohammadi

2021

Concurrency and Computation

View full text Add to dashboard Cite

Today, modern healthcare systems rely on advanced computational technologies, including cloud-based systems, to gather and examine personal health data on a large scale. The use of advanced cloud services technologies, such as software as a service, application as a service, is challenging for end-users of cloud systems to protect sensitive data in their health applications. According to the importance of publishing data in the cloud, the information should be recorded and handled somehow so that any individuals' identity remains hidden. Therefore, one of the critical privacy challenges is protecting the quality of published data and privacy-preserving on the healthcare cloud simultaneously. The K-anonymity technology is one of the prevalent methods used for privacy-preserving. In this article, we suggest a novel approach based on the clustering process using the K-means++ method to achieve an optimal k-anonymity algorithm. Also, we use the normal distribution function to delete data that is less frequent, which can be improved the quality of anonymized data. Extensive experiments show the proposed method has been able to reduce information loss 1.5 times and execution time 3.5 times compared to AKA and GCCG algorithms. Also, it is highly scalable than others.

show abstract

Optimization-based k-anonymity algorithms

Cited by 24 publications

References 30 publications

Privacy-preserving healthcare informatics: a review

Privacy-preserving healthcare informatics: a review

A Lightweight Approach for Origin-Destination Matrix Anonymization

A clustering‐based anonymization approach for privacy‐preserving in the healthcare cloud

Contact Info

Product

Resources

About