Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userPrivacy Preserving Techniques Applied to CPNI Data: Analysis and Recommendations
93
0
0.0
(
0
)
Added by
Jeffrey Murray Jr
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
With mobile phone penetration rates reaching 90%, Consumer Proprietary Network Information (CPNI) can offer extremely valuable information to different sectors, including policymakers. Indeed, as part of CPNI, Call Detail Records have been successfully used to provide real-time traffic information, to improve our understanding of the dynamics of peoples mobility and so to allow prevention and measures in fighting infectious diseases, and to offer population statistics. While there is no doubt of the usefulness of CPNI data, privacy concerns regarding sharing individuals data have prevented it from being used to its full potential. Traditional de-anonymization measures, such as pseudonymization and standard de-identification, have been shown to be insufficient to protect privacy. This has been specifically shown on mobile phone datasets. As an example, researchers have shown that with only four data points of approximate place and time information of a user, 95% of users could be re-identified in a dataset of 1.5 million mobile phone users. In this landscape paper, we will discuss the state-of-the-art anonymization techniques and their shortcomings.
rate research
Read More
In recent years, the data mining techniques have met a serious challenge due to the increased concerning and worries of the privacy, that is, protecting the privacy of the critical and sensitive data. Different techniques and algorithms have been already presented for Privacy Preserving data mining, which could be classified in three common approaches: Data modification approach, Data sanitization approach and Secure Multi-party Computation approach. This paper presents a Data modification- based Framework for classification and evaluation of the privacy preserving data mining techniques. Based on our framework the techniques are divided into two major groups, namely perturbation approach and anonymization approach. Also in proposed framework, eight functional criteria will be used to analyze and analogically assessment of the techniques in these two major groups. The proposed framework provides a good basis for more accurate comparison of the given techniques to privacy preserving data mining. In addition, this framework allows recognizing the overlapping amount for different approaches and identifying modern approaches in this field.
In the big data era, more and more cloud-based data-driven applications are developed that leverage individual data to provide certain valuable services (the utilities). On the other hand, since the same set of individual data could be utilized to infer the individuals certain sensitive information, it creates new channels to snoop the individuals privacy. Hence it is of great importance to develop techniques that enable the data owners to release privatized data, that can still be utilized for certain premised intended purpose. Existing data releasing approaches, however, are either privacy-emphasized (no consideration on utility) or utility-driven (no guarantees on privacy). In this work, we propose a two-step perturbation-based utility-aware privacy-preserving data releasing framework. First, certain predefined privacy and utility problems are learned from the public domain data (background knowledge). Later, our approach leverages the learned knowledge to precisely perturb the data owners data into privatized data that can be successfully utilized for certain intended purpose (learning to succeed), without jeopardizing certain predefined privacy (training to fail). Extensive experiments have been conducted on Human Activity Recognition, Census Income and Bank Marketing datasets to demonstrate the effectiveness and practicality of our framework.
Bitcoin and many other similar Cryptocurrencies have been in existence for over a decade, prominently focusing on decentralized, pseudo-anonymous ledger-based transactions. Many protocol improvements and changes have resulted in new variants of Cryptocurrencies that are known for their peculiar characteristics. For instance, Storjcoin is a Proof-of-Storage-based Cryptocurrency that incentivizes its peers based on the amount of storage owned by them. Cryptocurrencies like Monero strive for user privacy by using privacy-centric cryptographic algorithms. While Cryptocurrencies strive to maintain peer transparency by making the transactions and the entire ledger public, user privacy is compromised at times. Monero and many other privacy-centric Cryptocurrencies have significantly improved from the original Bitcoin protocol after several problems were found in the protocol. Most of these deficiencies were related to the privacy of users. Even though Bitcoin claims to have pseudo-anonymous user identities, many attacks have managed to successfully de-anonymize users. In this paper, we present some well-known attacks and analysis techniques that have compromised the privacy of Bitcoin and many other similar Cryptocurrencies. We also analyze and study different privacy-preserving algorithms and the problems these algorithms manage to solve. Lastly, we touch upon the ethics, impact, legality, and acceptance of imposing these privacy algorithms.
As machine learning becomes a practice and commodity, numerous cloud-based services and frameworks are provided to help customers develop and deploy machine learning applications. While it is prevalent to outsource model training and serving tasks in the cloud, it is important to protect the privacy of sensitive samples in the training dataset and prevent information leakage to untrusted third parties. Past work have shown that a malicious machine learning service provider or end user can easily extract critical information about the training samples, from the model parameters or even just model outputs. In this paper, we propose a novel and generic methodology to preserve the privacy of training data in machine learning applications. Specifically we introduce an obfuscate function and apply it to the training data before feeding them to the model training task. This function adds random noise to existing samples, or augments the dataset with new samples. By doing so sensitive information about the properties of individual samples, or statistical properties of a group of samples, is hidden. Meanwhile the model trained from the obfuscated dataset can still achieve high accuracy. With this approach, the customers can safely disclose the data or models to third-party providers or end users without the need to worry about data privacy. Our experiments show that this approach can effective defeat four existing types of machine learning privacy attacks at negligible accuracy cost.
Releasing full data records is one of the most challenging problems in data privacy. On the one hand, many of the popular techniques such as data de-identification are problematic because of their dependence on the background knowledge of adversaries. On the other hand, rigorous methods such as the exponential mechanism for differential privacy are often computationally impractical to use for releasing high dimensional data or cannot preserve high utility of original data due to their extensive data perturbation. This paper presents a criterion called plausible deniability that provides a formal privacy guarantee, notably for releasing sensitive datasets: an output record can be released only if a certain amount of input records are indistinguishable, up to a privacy parameter. This notion does not depend on the background knowledge of an adversary. Also, it can efficiently be checked by privacy tests. We present mechanisms to generate synthetic datasets with similar statistical properties to the input data and the same format. We study this technique both theoretically and experimentally. A key theoretical result shows that, with proper randomization, the plausible deniability mechanism generates differentially private synthetic data. We demonstrate the efficiency of this generative technique on a large dataset; it is shown to preserve the utility of original data with respect to various statistical analysis and machine learning measures.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
