Privacy Preserving Distributed Data Mining Research Articles

A Reliable Application of MPC for Securing the Tri-Training Algorithm

Due to the widespread use of distributed data mining techniques in a variety of areas, the issue of protecting the privacy of sensitive data has received increasing attention in recent years. Privacy-preserving distributed data mining (PPDDM) focuses on decentralized data analysis without the disclosure of sensitive information from data owner. However, the previous PPDDM mostly works on a limited amount of labeled data. In contrast to the real world, unlabeled data is abundance and labeled data is scarce. The objectives of this paper are to study and to analyze privacy-preserving properties of semi-supervised learning (SSL) algorithm with the combination of labeled and unlabeled data, where data is distributed among multiple data owners. In this paper we propose a Privacy-preserving Distributed Data Mining (PPDDM) method by designing a reliable application of secure MPC to semi-supervised tri-training algorithms. We simulate the original tri-training algorithm and tri-training algorithm with secure MPC using a different types of classifiers and datasets. The simulation results show that tri-training in secure MPC has almost same accuracy compared to original tri-training algorithm. We also compare execution time in addition to performance evaluation of tri-training in secure and the original tri-training algorithms.

A systematic review on privacy-preserving distributed data mining

Combining and analysing sensitive data from multiple sources offers considerable potential for knowledge discovery. However, there are a number of issues that pose problems for such analyses, including technical barriers, privacy restrictions, security concerns, and trust issues. Privacy-preserving distributed data mining techniques (PPDDM) aim to overcome these challenges by extracting knowledge from partitioned data while minimizing the release of sensitive information. This paper reports the results and findings of a systematic review of PPDDM techniques from 231 scientific articles published in the past 20 years. We summarize the state of the art, compare the problems they address, and identify the outstanding challenges in the field. This review identifies the consequence of the lack of standard criteria to evaluate new PPDDM methods and proposes comprehensive evaluation criteria with 10 key factors. We discuss the ambiguous definitions of privacy and confusion between privacy and security in the field, and provide suggestions of how to make a clear and applicable privacy description for new PPDDM techniques. The findings from our review enhance the understanding of the challenges of applying theoretical PPDDM methods to real-life use cases, and the importance of involving legal-ethical and social experts in implementing PPDDM methods. This comprehensive review will serve as a helpful guide to past research and future opportunities in the area of PPDDM.

Privacy preserving distributed data mining based on secure multi-party computation

Data mining is an important task to understand the valuable information for making correct decisions. Technologies for mining self-owned data of a party are rather mature. However, how to perform distributed data mining to obtain information from data owned by multiple parties without privacy leakage remains a big challenge. While secure multi-party computation (MPC) may potentially address this challenge, several issues have to be overcome for practical realizations. In this paper, we point out two unsupported tasks of MPC that are common in the real-world. Towards this end, we design algorithms based on optimized matrix computation with one-hot encoding and LU decomposition to support these requirements in the MPC context. In addition, we implement them based on a SPDZ protocol, a computation framework of MPC. The experimental evaluation results show that our design and implementation are feasible and effective for privacy preserving distributed data mining.

Mixture-Model-Based Graph for Privacy-Preserving Semi-Supervised Learning

Privacy has become a major concern in data mining as it is utilized in many important applications. Distributed privacy-preserving data mining (DPPDM) is one of the techniques to address this concern, which focuses on protecting private information of members in distributed systems during data mining. As DPPDM is widely discussed in recent works, the semi-supervised manner of learning still draws less attention in this field. In this paper, a mixture-model-based semi-supervised DPPDM method is proposed. By introducing our method, a site in a distributed system is able to initiate a learning process using labeled data of its own and unlabeled data from all the sites. During the process, no individual data of any site is revealed to others, no information about data can be traced back to any specific site, and only the initiating site learns the result. We propose a parameter-masking privacy-preserving Expectation-Maximization (EM) algorithm and a mixture-model-based semi-supervised learning algorithm as the two main steps of our method. Experiments on both synthetic and real-world data demonstrate the effectiveness of the proposed method.

A sanitization approach for privacy preserving data mining on social distributed environment

Data owners worry about their private data in the information that is being uncovered without authorization in the cloud computing environment. While applying privacy preserving methods to the data, the data owners attempt to retain the knowledge inside the data. One approach to solve this problem is the concept of distributed databases where different parties have horizontal or vertical partitions of the data. Cluster analysis is a frequently used data mining task which aims at decomposing or partitioning a usually multivariate data set into groups such that the data objects in one group are more similar to each other. While using encryption based kernel k-means algorithm, large data’s can’t be encrypted in the distributed environment. To extend the privacy concept, a novel method based Privacy Preserving Distributed Data Mining is planned. According to this, a sanitization approach will be developed to improve the privacy of the user data. In sanitization process, a privacy based objective function will be developed and an optimal key will be generated based on the proposed objective function. Here artificial bee colony algorithm will be utilized for optimal key generation and large amount of data can be encrypted. Once the sanitization process is done, the sanitized information will be updated to service provider by the helper user for each cluster. Finally, the experimentation will be carried out with existing database to prove the efficiency of the proposed algorithm. The implementation will be done in JAVA using cloud simulator. Extensive execution assessments and security analysis exhibit the legitimacy and efficiency of the proposed technique.

A New Method to Compute Ratio of Secure Summations and Its Application in Privacy Preserving Distributed Data Mining

Computing the ratio of secure summations (RSS) is one of the most important tools for privacy preserving distributed data mining. It refers to such a problem; given the n parties and their respective secret values (x i ,y i ), where i=1, ..., n , how can they get the value of Σ i=1 n x i / Σ i=1 n y i without disclosing (x i ,y i ), and Σ i=1 n x i and Σ i=1 n y i In this paper, we propose a new method to solve this problem. Compared with the existing methods, our method is not only secure under the assumption of semi-honest but also can resist collusion attacks-even when all but one party are corrupt. At the same time, the computational complexity and communication complexity of our method are both O(n) . Therefore, it can meet the needs of the practical application in terms of both security and efficiency. In addition to the theoretical guarantee, the practical usability of our method is also verified through experiments by constructing Naive Bayes classifiers in a privacy-preserving distributed environment.

Privacy-Preserving Distributed Data Mining Techniques: A Survey

distributed data mining settings, leakage of the real data is not adequate because of privacy issues. To overcome this problem, numerous privacy-preserving distributed data mining practices have been suggested such as protect privacy of their data by perturbing it with a randomization algorithm and using cryptographic techniques. In this paper, we review and provide extensive survey on different privacy preserving data mining methods and analyses the representative techniques for privacy preserving data mining. We majorly discuss the distributed privacy preservation techniques which provide secure solutions using primitive operations of cryptographic protocols such as secure multi-party computation (SMPC), secret sharing schemes (SSS) and homomorphic encryption (HC). Keywordsmining, K-means clustering, Data privacy, Privacy preserving, Multiparty computation, Threshold Cryptography.

Study on distributed privacy preserving data mining

Recently, with the rapid development of network, communications and computer technology, privacy preserving data mining (PPDM) has become an increasingly important research in the field of data mining. In distributed environment, how to protect data privacy while doing data mining jobs from a large number of distributed data is more far-researching. This paper describes current research of PPDM at home and abroad. Then it puts emphasis on classifying the typical uses and algorithms of PPDM in distributed environment, and summarizing their advantages and disadvantages. Furthermore, it points out the future research directions in the field.

A generic and distributed privacy preserving classification method with a worst-case privacy guarantee

It is often necessary for organizations to perform data mining tasks collaboratively without giving up their own data. This necessity has led to the development of privacy preserving distributed data mining. Several protocols exist which deal with data mining methods in a distributed scenario but most of these methods handle a single data mining task. Therefore, if the participating parties are interested in more than one classification methods they will have to go through a series of distributed protocols every time, thus increasing the overhead substantially. A second significant drawback with existing methods is that they are often quite expensive due to the use of encryption operations. In this paper a method has been proposed that addresses both these issues and provides a generic approach to efficient privacy preserving classification analysis in a distributed setting with a worst-case privacy guarantee. The experimental results demonstrate the effectiveness of this method.

Distributed RK- Secure Sum Protocol for Privacy Preserving

Secure multi party computation allows several parties to compute some function of their inputs without disclosing the actual input to one another. Secure sum computation is an easily understood example and the component of the various secure multi party computation solutions. Secure sum computation allows parties to compute the sum of their individual inputs without disclosing the inputs to one another. In this paper, we propose a protocol with more security, when a group of the computing parties want to know the data of some other party. I. INTRODUCTION The huge growth of the Internet and its easy access by common man created opportunities for joint computations by multiple parties. All the participating parties for the sake of their mutual benefit want to compute the common function of their inputs but at the same time they are worried about the privacy of their data. This subject of the information security is called secure multi party computation. This subject has two goals; one is the privacy of the individual data inputs and another is the correctness of the result. Mainly two models exist in the literature for the analysis of the secure multi party computation problems. Ideal model of the secure multi party computation uses a Trusted Third Party apart from the participating parties. Parties supply their inputs to the Trusted Third Party. Computation of the function is done by the Trusted Third Party and then the result is sent to all the parties. In this paradigm the trustworthiness of the Trusted Third Party is critically important because if the Trusted Third Party turns corrupt, it can supply the private inputs of one party to others. But it is extensively used model of the secure multi party computation due to its easy implementation and the protocols available which prevent the Trusted Third Party to act maliciously. Real model of the secure multi party computation does not use any Trusted Third Party but the parties themselves agree on some protocol for the computation. The parties behavior in the secure multi party computation is important to consider. An honest party follows the protocol and respects the privacy of other parties. A semi honest party follows the protocol but also tries to learn other information than the result. The corrupt party neither follows the protocol nor respects the privacy of other parties. Different protocols are needed for different secure multi party computation models and the behavior of the party. Solutions are available for secure multi party computation problems using cryptographic techniques, randomization techniques and anonymization methods. The subject of secure multi party computation has been evolved from two party comparison problems (1) to multiparty image template matching problems. Many specific secure multi party computation problems have been defined and analyzed by researchers like Private Information Retrieval, Selective Function Evaluation, Privacy-Preserving Database Query, Privacy-Preserving Geometric Computation, Privacy- Preserving Statistical Analysis, Privacy-Preserving Intrusion Detection and Privacy-Preserving Cooperative Scientific Computation. Based on these general secure multi party computation problems many real life applications have been emerged like Privacy- Preserving Electronic Voting, Privacy-Preserving Bidding and Auctions, Privacy-Preserving Social Network Analysis, Privacy-Preserving Signature and Face Detection, Etc. Secure sum computation problem of secure multi party computation can be defined as: How multiple parties can compute the sum of their input values without disclosing actual values to one another. Secure sum can work as the tool for the secure multi party computation solutions in the privacy preserving distributed data mining problems (2). We proposed a secure sum protocol using random numbers (2). In this paper, we proposed a novel changing neighbors approach Distributed RK secure sum protocol for achieving more security in case a group of the parties collude to know the private data of some other party. 1.1 Distributed Database Partition: - Database is divided into three main portioning horizontal partition, vertical partition and hybrid partition. 1.1.1 Horizontal Partitioning: - Horizontal partitioning divides the whole database into the number of small database according to the row splitting. So that the execution of query will be very fast as well as we will be able to provide more privacy to the portioned database.

Privacy preserving distributed knowledge discovery: survey and future directions

The aim of privacy preserving data mining algorithms is to extract relevant knowledge from large amounts of data while protecting at the same time sensitive information. In the first part of this paper, we discuss privacy preserving distributed data mining techniques and provide a survey on the state–of–the–art methods in this field. We have considered randomisation, k–anonymisation, and cryptographic techniques. In the second part of the paper, we have described our proposed system for privacy preserving knowledge discovery over distributed databases which is still under development phase. The system is designed to perform local operations (local mining) in each site. This produces intermediate data that can be used to obtain the final result without revealing the private information of any site. Our proposal is mainly based on association rule mining and cryptographic techniques.

Communication-Efficient Privacy-Preserving Clustering

The ability to store vast quantities of data and the emergence of high speed networking have led to intense interest in distributed data mining. However, privacy concerns, as well as regulations, often prevent the sharing of data between multiple parties. Privacy-preserving distributed data mining allows the cooperative computation of data mining algorithms without requiring the participating organizations to reveal their individual data items to each other. This paper makes several contributions. First, we present a simple, deterministic, I/O-efficient kclustering algorithm that was designed with the goal of enabling an efficient privacy-preserving version of the algorithm. Our algorithm examines each item in the database only once and uses only sequential access to the data. Our experiments show that this algorithm produces cluster centers that are, on average, more accurate than the ones produced by the well known iterative k-means algorithm, and compares well against BIRCH. Second, we present a distributed privacy-preserving protocol for k-clustering based on our new clustering algorithm. The protocol applies to databases that are horizontally partitioned between two parties. The participants of the protocol learn only the final cluster centers on completion of the protocol. Unlike most of the earlier results in privacy-preserving clustering, our protocol does not reveal intermediate candidate cluster centers. The protocol is also efficient in terms of communication and does not depend on the size of the database. Although there have been other clustering algorithms that improve on the k-means algorithm, ours is the first for which a communication efficient cryptographic privacy-preserving protocol has been demonstrated.

Privacy-preserving data mining in the malicious model

Most of the cryptographic work in privacy-preserving distributed data mining deals with semi-honest adversaries, which are assumed to follow the prescribed protocol but try to infer private information using the messages they receive during the protocol. Although the semi-honest model is reasonable in some cases, it is unrealistic to assume that adversaries will always follow the protocols exactly. In particular, malicious adversaries could deviate arbitrarily from their prescribed protocols. Secure protocols that are developed against malicious adversaries require utilisation of complex techniques. Clearly, protocols that can withstand malicious adversaries provide more security. However, there is an obvious trade-off: protocols that are secure against malicious adversaries are generally more expensive than those secure against semi-honest adversaries only. In this paper, our goal is to make an analysis of trade-offs between performance and security in privacy-preserving distributed data mining algorithms in the two models. In order to make a realistic comparison, we enhance commonly used subprotocols that are secure in the semi-honest model with zero knowledge proofs to be secure in the malicious model. We compare the performance of these protocols in both models.

Privacy-preserving Naïve Bayes classification

Privacy-preserving data mining--developing models without seeing the data --- is receiving growing attention. This paper assumes a privacy-preserving distributed data mining scenario: data sources collaborate to develop a global model, but must not disclose their data to others. The problem of secure distributed classification is an important one. In many situations, data is split between multiple organizations. These organizations may want to utilize all of the data to create more accurate predictive models while revealing neither their training data/databases nor the instances to be classified. Naive Bayes is often used as a baseline classifier, consistently providing reasonable classification performance. This paper brings privacy-preservation to that baseline, presenting protocols to develop a Naive Bayes classifier on both vertically as well as horizontally partitioned data.

Random projection-based multiplicative data perturbation for privacy preserving distributed data mining

This paper explores the possibility of using multiplicative random projection matrices for privacy preserving distributed data mining. It specifically considers the problem of computing statistical aggregates like the inner product matrix, correlation coefficient matrix, and Euclidean distance matrix from distributed privacy sensitive data possibly owned by multiple parties. This class of problems is directly related to many other data-mining problems such as clustering, principal component analysis, and classification. This paper makes primary contributions on two different grounds. First, it explores independent component analysis as a possible tool for breaching privacy in deterministic multiplicative perturbation-based models such as random orthogonal transformation and random rotation. Then, it proposes an approximate random projection-based technique to improve the level of privacy protection while still preserving certain statistical characteristics of the data. The paper presents extensive theoretical analysis and experimental results. Experiments demonstrate that the proposed technique is effective and can be successfully used for different types of privacy-preserving data mining applications.

Tools for privacy preserving distributed data mining

Privacy preserving mining of distributed data has numerous applications. Each application poses different constraints: What is meant by privacy, what are the desired results, how is the data distributed, what are the constraints on collaboration and cooperative computing, etc. We suggest that the solution to this is a toolkit of components that can be combined for specific privacy-preserving data mining applications. This paper presents some components of such a toolkit, and shows how they can be used to solve several privacy-preserving data mining problems.