Privacy-preserving data publishing: A survey of recent developments

By Benjamin C. M. Fung, Ke Wang, Rui Chen, Philip S. Yu. 2010

1,918 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Sociology and Political Science, Privacy & Security

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Sociology and Political Science

In Privacy & Security

Citations in the corpus

A Comprehensive Comparison of Multiparty Secure Additions with Differential Privacy (2015, background): “Traditional approaches such as removing identifying attributes, generalizing, or perturbing individual attribute values, are susceptible to various attacks [26].”
Signal Processing and Machine Learning with Differential Privacy (2013, background): “There are many definitions and models for privacypreserving computation, and a recent survey by Fung et al. compares several different approaches [1].”
Learning privately from multiparty data (2016, background): “To preserve privacy in data publishing, several approaches such as k-anonymity (Sweeney, 2002) and secure multiparty computation (Yao, 1982) have been proposed (see (Fung et al., 2010) for a review.”
Differential Privacy Techniques for Cyber Physical Systems: A Survey (2018, methodology): “Anonymization techniques initiated from kanonymity, to l-diversity, and then to t-closeness method [54].”
AI in Medical Imaging Informatics: Current Challenges and Future Directions (2020, neutral): “PPDP objective is to preserve patients’ privacy while achieving the minimum possible loss of information [53].”
Privacy in the Internet of Things: threats and challenges (2014, background): “Identity protection and, complementary, protection against identiﬁcation is a predominant topic in RFID privacy (Section 3.1), but has also gained much attention in the areas of data anonymization [73, 74, 75], and privacy-enhancing identity management [76, 77].”
Secure multiparty aggregation with differential privacy: a comparative study (2013, background): “Traditional approaches such as removing identifying attributes, generalizing, or perturbing individual attribute values, are susceptible to various attacks [21].”
Pseudonym Schemes in Vehicular Networks: A Survey (2015, background): “[184] for details on k-anonymity, and its optimizations l-diversity and t-closeness.”
Computing with Spatial Trajectories (2011, background): “Several studies have analyzed and evaluated various line generalization algorithms mathematically and perceptually and ranked the DouglasPeucker algorithm highly [18, 32, 25].”
Towards an axiomatization of statistical privacy and utility (2010, background): “For surveys on statistical privacy, see [10, 21 , 1]. The need for a better understanding of privacy denitions and privacy mechanisms was underscored by the work of Dinur and Nissim [12] (and later by Dwork et al. [17]) that showed that some intuitive methods for preserving privacy actually did not”
Differentially private data cubes: optimizing noise sources and consistency (2011, background): “Compared to previous techniques for privacy-preserving data publishing (see [2, 15] for surveys), DP makes very conservative assumptions about the adversary’s background knowledge.”
Discrimination-aware data mining: a survey (2017, background): “Fung et al. (2010) have presented a detailed survey in this area.”
Differential Privacy via Wavelet Transforms (2009, background): “A plethora of techniques have been proposed for privacy-preserving data publishing (see [1], [2] for surveys).”
Pufferfish (2014, background): “Because of this, the resulting privacy deﬁnitions tend to be less secure and are often subject to new kinds of attacks (see the surveys [Chen et al. 2009; Fung et al. 2010; Adam and Worthmann 1989] for more details). is also often not clear what data protecting individual records.”
SplitX: high-performance private analytics (2013, background): “There is a massive literature on database privacy that assumes a trusted database, including [11,22,31] (see [16] for a survey).”
Technical Privacy Metrics (2015, background): “ max), where 8y2P: minPr(y2D TjD P) max However, this model assumes that the data publisher and the adversary have access to the same external tables. This assumption may not hold in practice [Fung et al. 2010]. 5.5.5Hiding Property In communication systems, the source (or destination) hiding prop”
Generation of Synthetic Electronic Medical Record Text (2018, methodology): “The anonymization is typically done by applying generalization and suppression operations to modify the patients’ attributes [1].”