An Efficient Algorithm for Privacy Preserving Maximal Frequent Itemsets Mining

Privacy Preservation in Data Mining (PPDM) including for Privacy Preserving Association Rule Mining (PPARM) has attracted lots of attention in recent research and practice. However, the current method or approach still have drawbacks in the sense that there are trade-offs between efficiency and privacy preservation. This paper describes our work towards providing a new efficient PPARM protocol. We reviewed current literature on PPARM and mapped the methods or approaches involved. As previous research showed that Elliptic Curve Cryptography (ECC) perform better than the other Public Key systems such as RSA and Diffie-Hellman, we will utilize ECC for reducing the computational cost of the new PPARM protocol. In choosing good elliptic curves for ECC, we measured the running time of the key generation for various group of recommended elliptic curves i.e. Brainpool curves (by Brainpool), Prime, C2pnb, C2tnb curves (by ANSI X9.62), Secp curves (by SECG), and PrimeCurve curves (by CDC Group). As the result, Secp curves outperformed all of the other curves on overall average ratio of running time and key size of key generation by 4.4% up to 357.6%.

A number of privacy preserving techniques have been proposed recently in data mining. In this paper, we provide a review of the state-of-the-art methods for privacy preserving data mining. and discuss methods for randomization, secure multipart computation, and so on. We also make a classification for the privacy preserving data mining technologies, and analyze some works in this field, such as data distortion method for achieving privacy preserving association rule mining. Detailed evaluation criteria of privacy preserving algorithm were illustrated, which include algorithm performance, data utility, and privacy protection degree. Finally, the development of privacy preserving data mining for further directions is given.

Association rules discovered by association rule mining may contain some sensitive rules, which may cause potential threats towards privacy and security. In this paper, we address the problem of privacy preserving association rule mining by proposing a Knowledge Sanitization to prevent the disclosure of sensitive rules. We also present parallel approach to generate a non-sensitive and sensitive rule mining. Our proposed solution key issue is to provide a high degree of parallelism for privacy to the data owner.

In the discipline of data mining, association rule mining is an important study topic that focuses on discovering the relationships between database attributes. The maximum frequent itemset comprises the information of all frequent itemsets, which is one of the important difficulties in mining association rules, and certain data mining applications just need to mine the maximum frequent itemsets. As a result, analyzing the maximum frequent itemset mining technique is practical. Considering this, the research introduces FP-MFIA, a new maximum frequent itemset mining approach based on the FP-tree, which is inspired by the data structure of the frequent pattern tree and the idea that the maximum frequent itemset implies all frequent itemsets. First, the FP-MFIA constructs a one-way FP-tree structure, which only has pointers from the root to the leaves, so that only two scans of the FP-tree are required by the FP-MFIA. On the other hand, it redefines a data storage structure MFI-list for maximum frequent itemsets. It can quickly release unnecessary nodes in the FP-tree after scanning it. In this way, not only the information required by the maximum frequent itemsets can be quickly mined but also the space required for storing the maximum frequent itemsets can be reduced, which greatly improves the mining efficiency. Finally, experiments were conducted to compare the mining efficiency of the novel FP-MFIA algorithm to the IDMFIA and DMFIA algorithms. We can see from the findings that the FP-MFIA algorithm is more efficient than the other two techniques.

Maximal frequent item sets are one of several condensed representations of frequent item sets, which store most of the information contained in frequent item sets using less space, thus being more suitable for stream mining. This paper focuses on mining maximal frequent item sets approximately over a stream landmark model. We separate the continuously arriving transactions into sections, and the mining results are indexed by an extended direct update tree, thus, a simple but effective algorithm named SMIS is proposed. In our algorithm, we employ the Chern off Bound to perform the maximal frequent item set mining in a false negative manner, which can reduce the memory cost, as well guarantee our algorithm conducting with an incremental fashion. Our experimental results on two synthetic datasets and two real world datasets show that SMIS achieves much reduced memory cost in comparison with the state-of-the-art algorithm with a 100 percent precision.

In this research work, we design a security protocol that derives association rules securely from the horizontally distributed databases without a trusted third party (TTP), even communication channel is unsecured between involving sites. It ensures privacy and security of owner’s with the help of elliptic curve based Diffie-Hellman and Digital Signature Algorithm.

High-performance implementation of evolutionary privacy-preserving algorithm for big data using GPU platform

The recent advancement in data mining technology to analyze vast amount of data has played an important role in several areas of Business processing. Data mining also opens new threats to privacy and information security if not done or used properly. The main problem is that from non-sensitive data, one is able to infer sensitive information, including personal information, fact or even patterns which are generated by any algorithm of data mining. In order to focusing on privacy preserving association rule mining, the simplistic solution to address the problem of privacy is presented. The solution is to survey different aspects which are discussed in the several research papers and after analyzing those research papers conclude a new solution which is best in efficiency and performance. In this paper we propose a novel algorithm named Cloud Subset Preserving Mining (CSPM). The entire system architecture consists of three phases: 1) Check for Authentication. 2) Reading the database. 3) Perform Pruning. Our algorithm is a good way to apply data mining techniques with security that hides our logical instances from others. The all the operations are performed in cloud computing environment.

Frequent itemsets (FIs) mining from uncertain database is a very popular research area nowadays. Many algorithms have been proposed to mine FI from uncertain database. But in typical FI mining process, all the FIs have to be mined individually, which needs a huge memory. Four trees are proposed in this paper which are: (i) maximal frequent itemset from uncertain database (MFU) tree which contains only the maximal frequent itemsets generated from uncertain database, (ii) closed frequent itemset from uncertain database (CFU) tree which contains only closed frequent itemsets generated from uncertain database, (iii) maximal frequent itemset from uncertain data stream (MFUS) tree which contains maximal frequent itemsets generated from uncertain data stream and (iv) closed frequent itemset from uncertain data stream (CFUS) tree which contains closed frequent itemsets generated from uncertain data stream. Experimental results are also presented which show that maximal and closed frequent itemsets mining requires less time and memory than typical frequent itemsets mining.

Privacy preserving association rule mining has been an active research area since recently. To this problem, there have been two different approaches--perturbation based and secure multiparty computation based. One drawback of the perturbation based approach is that it cannot always fully preserve individual's privacy while achieving precision of mining results. The secure multiparty computation based approach works only for distributed environment and needs sophisticated protocols, which constrains its practical usage. In this paper, we propose a new approach for preserving privacy in association rule mining. The main idea is to use keyed Bloom filters to represent transactions as well as data items. The proposed approach can fully preserve privacy while maintaining the precision of mining results. The tradeoff between mining precision and storage requirement is investigated. We also propose ?-folding technique to further reduce the storage requirement without sacrificing mining precision and running time.

Objective:Privacy protection in the medical field means the protection of individuals from being associated with undesirable conditions, diagnoses or treatments (Sensitive Attributes). The problem of knowledge discovery from health care data by applying data mining algorithms is inversely related to the privacy of individuals. Due to the tremendous growth of data in a large scale, there is a demand to protect the sensitive data accessible from medical datasets.Methods:This paper considers the problem of building privacy preserving association rule mining algorithm using the notion of TF * IDF derived from the information retrieval domain. The highly sensitive transaction is chosen using the product of Relative Item Frequency and Condensed Frequency. Finally, sensitive fuzzy data is perturbed to hide these refined rules.Results:It has been found that the number of non-sensitive rules lost as a side effect of hiding sensitive rule is 20% less and number of ghost rules is 30% less in proposed work than in previous work using Transactional Impact factor method. The execution time of hiding a rule is 26% lesser on an average in the proposed technique for various values of minimum confidence threshold. It has been observed that the number of modifications to the original dataset after hiding three rules were reduced by 66% in proposed method than in previous work. As the number of modifications to original data is less the chances of generating false association is also reduced.Conclusion:In this paper, a novel method was presented to hide the sensitive rule in quantitative data by decreasing the support of the RHS of the rule. Experimental results demonstrate that the proposed approach is more efficient as it facilitates better rule hiding and minimizes the number of lost rules and ghost rules. Also, this approach makes minimum modifications to the dataset.

Privacy considerations often constrain data mining projects. This paper addresses the problem of association rule mining where transactions are distributed across sources. Each site holds some attributes of each transaction, and the sites wish to collaborate to identify globally valid association rules. However, the sites must not reveal individual transaction data. We present a two-party algorithm for efficiently discovering frequent itemsets with minimum support levels, without either site revealing individual transaction values.

Privacy-preserving data mining aims at discovering beneficial information from a large amount of data without violating the privacy policy. Privacy-preserving association rules mining research has already generated many interesting results. Based on commutative encryptions and the Secure Multi-party Computation (SMC) theory, Kantarcioglu and Clifton [1] propose two protocols to implement privacy-preserving mining of association rules over horizontally partitioned data. The paper addresses its incorrect security proof and introduces a more well-founded proof. This paper also identifies several other errors in [1]. This kind of protocols and their proof are a concrete application of Secure Multi-party Computation, which is be of great significances to the privacy-preserving data mining studies based on SMC. Thus establishment of the correct proof methodology is important. This paper demonstrates the correct proof methodology by correcting the fault proof in [1]

Privacy-Preserving Association Rule Mining Considering Multi-objective through an Evolutionary Algorithm

The research in this paper analyzes the origin of the data generated and treated as Big Data and discusses the hypothesis about the conflict generated by the management of personal data privacy. It is based on a model to identify the set of management practices and resources, resulting from the integration of guidelines, norms, standards and contractual commitments that are more relevant to keeping personal data safe and generating trust in the environment. The field study shows the importance and urgency of this issue, given that it is necessary to innovate in business management, the provision of public services and the design and implementation of regional development policies. Through its conclusions, it is verified that an intelligent management is needed to rectify the course in terms of discovering and detecting patterns, relations and formulating models from these gigantic databases. This work presents a generic model for the management of privacy in the cloud of huge volumes of data. The impact of privacy is analyzed, risks are identified, and solutions provided by standards are explored to develop controls that can be integrated into a model to establish the steps so that any type of public or private organization can verify the organizational impact of its products, processes or service in the fulfillment of its strategic objectives.KeywordsBig DataStandardsDatabase and large data managementData modelsData privacySecurity