Privacy Preserving For Outsourced Data
issue 1

Privacy Preserving For Outsourced Data

K. S Chandwani1, Anjali Mishra2, Ayushi Desai3, Priyanka Kushwaha4, Sonal Nikose5, Shivani Wanjari6
Professor, Department of Computer Technology, KDK College of Engineering, Nagpur, Maharashtra, India
2-6CSE, Smt.. Rajshri Mulak College of Engineering for Women, Nagpur, Maharashtra, India


Privacy-preserving data mining is one of the newest trends in privacy and security research. It is determined by one of the significant arrangement issues of the information era – the right to privacy. With the use of cloud computing services, an organization lack in ability or computational resources can outsource its mining needs to an outsider service provider. However, both the items and the association rules of the outsourced database are viewed as private property of the enterprise. To maintain privacy, the data owner transforms its data to the server then sends mining queries to the server and recovers the real pattern from the extracted patterns received from the outsider server. In this system studied the issue of outsourcing the association rule mining framework within a corporate privacy-preserving framework. The Rob Frugal technique is introduced with defeat the privacy vulnerabilities of outsourced information. It is an encryption plan, based on one to one substitution ciphers for items and including fake patterns for database. Here the attackers/hackers can discover information by guessing attack, also man in the middle attack is possible on Rob Frugal encryption to conquer this issue, the proposed procedure encompasses Paillier encryption so as to enhance the security level for outsourced information with less complexity and to protect against the forging the contents of the correspondence. The principle thought is to preserve privacy of transactional data at the Server side.

Index Terms– Access Control, Association rules mining, FP-Growth algorithm, Outsourced Databases, Paillier Encryption.


“Protection Preservation” in information mining implies the Confidential or critical information must be jelly or secure by the unapproved individual or assailant. Privacy Preserving Data Mining (PPDM) is used to extract useful knowledge from large amount of data and at the same time protect the sensitive data from the data miners.The issue of security protecting information mining has turned out to be more vital as of late due to the expanding capacity to store individual information about clients, and corporate information of private establishment with the end goal of outsourcing and a wide range of different purposes. In any case, when the all of information be put in outsourced database administration supplier, the supplier is not trusted, touchy information may have spilled emergency. Amazon Dynamo DB, Hosted MongoDB is a few cases of database administration suppliers. Protecting the security of the outsourced databases is an extraordinary test in the current scenario. As the information is put away at the administration provider’s site, the facts may confirm that administration supplier is sceptical as far as uncovering and abusing the information. For this situation, security of the database can be hampered significantly. In the event that appropriate security is not authorized, then there are odds of information ruptures and hacking the information in an unapproved way. Information breaking implies unveiling the delicate information purposefully or inadvertently. As indicated by the information disagreement examination done by Trust wave in 2012, 76% of security absence were created by the unknown administration supplier. Along these lines, it is extremely fundamental for the organizations to know about security completing in their outsourced databases to keep the information private and in this manner following the administration standards and controls. Thus, actualizing them in an efficient way is critical from the security perspective. Different strategies are utilized for understanding the security as a part of database outsourcing. These systems incorporate encryption, verified information structures, request safeguarding encryption, signature plans, and so forth. In this paper, we have given the complete investigation of security strategies alongside their advantages and disadvantages. The target of this paper is to concentrate for the most part on different security procedures for outsourced exchange datasets. The rest of the part of the paper is composed like this Section II shows the hypothetical foundation of this paper. Area III presents similar study/investigation of various security strategies and segment IV closes the paper with outline and future heading.


The diagram shows the interaction between the user and server. Here user sends his/her file over the cloud, but before that the file will be encrypted by using two encryption algorithms such as Rob Frugal and Paillier. The file firstly encrypted using Rob frugal which follows some steps like One to One substitution, Support Calculation, Frugal Grouping, Robust k-Grouping method (Rob Frugal Grouping), Fake Transaction Construction, which are described as below:

Fig. Flow of Method

1 to 1 substitution cipher: The method which transformed original transaction database D into its encrypted version D*. Table 1(a) shows original transaction while Table 1(b) shows transaction after one to one substitution (encrypted).

Soda Nuts
Soda Milk
Milk Soda
Nuts Milk
Soda Dates
Nuts Soda
Soda Egg
Nuts Cake


e6 e5
e6 e4
e4 e6
e5 e4
e6 e2
e5 e6
e6 e3
e5 e1

Table1 (b): TDB*

Support Calculation: Support count is the second step where number of time the items occurred in the original transaction database

Frugal Grouping:td


Table 2: Descending order of items based on their item support

Robust k-Grouping method: Where k be the group size (i.e. 2 or 3), Here we consider the group size as 2 .Given the items support table, from a group of size k such that no two items from any original transaction comes adjacent to each other i.e. we can’t group e6,e5 or e6,e2 as they occurs adjacent in original transaction. After K-grouping method we get output as:


Table 3: Rob Frugal with K-robust grouping

Fake Transaction Construction: We can construct Fake transaction by adding Noise in to original transaction i.e. we can add e1 4 times, similarly e3 3times and e2 2 times in original transaction, so we generate transaction from given noise as {e1, e3, e2}, {e1, e3}, {e1}. After that the file will be encrypted using the Paillier encryption. By giving some threshold value the file is send to server. At server side Association rule mining will be performed which will be done by the Apriori and FP growth algorithm. Now whenever user request for his/her file, then server will accepts the request and send the file to the user. Now the file will be decrypted only one time by using the Paillier decryption method, the file is still encrypted by the Rob frugal, and hence by this we are providing the security to file. By using Rob Frugal and Paillier encryption method we are able to select required data even in encrypted form.

Fig. Flowchart of Method


In this section we are going to discuss the comparison graph between the existing and proposed system.

Fig. Time Comparison Graph

The above graph shows the time comparison graph between the existing and proposed system. It illustrates that the time required for rule generation for both FP growth algorithm and Apriori algorithm. From the above illustrated graph we can conclude that the time required for FP-growth is less than the time required for Apriori algorithm. In this report, we surveyed our approaches towards addressing two specific problems: (i) privacy-preserving data management (PPDM) and (ii) privacy-preserving data analytics (PPDA). Also, we pointed out different issues that still need to be addressed to provide proper solutions to PPDM and PPDA. Addressing these issues will be the primary focus of our future work. A plain one-to-one substitution cipher is vulnerable to attacks; therefore, we use Paillier Homomorphic encryption algorithm which gives a better security than existing rob frugal algorithm. Through this paper we focus on encrypted data, other data transformation techniques such as anonymization can also be considered.


  1. R. Kaur, M. Sharma and S. Taruna, “Privacy Preserving Data Mining Model for the Social Networking,” 2019 6th International Conference on Computing for Sustainable Global Development (INDIACom), New Delhi, India, 2019, pp. 763-768.
  2. Yung-Wang Lin, Li-Cheng Yang, Luon-Chang Lin, and Yeong-Chin Chen, Preserving Privacy in Outsourced Database, International Journal of Computer and Communication Engineering, Vol. 3, No. 5, September 2015.
  3. Sumeet Bajaj, RaduSion, TrustedDB: A Trusted Hardware based Database with Privacy and Data Confidentiality, In Proc. of IEEE Transactions on Knowledge and Data Engineering, 2013.
  4. FoscaGiannotti, Laks V. S. Lakshmanan, Anna Monreale, Dino Pedreschi, and Hui (Wendy) Wang, Privacy-Preserving Mining of Association Rules From Outsourced Transaction. Databases, IEEE SYSTEMS JOURNAL, VOL. 7, NO. 3, SEPTEMBER 2012.
  5. Lena Wiese, Horizontal Fragmentation for Data Outsourcing with Formula-Based Confidentiality Constraints, Advance in information and computer security, Springer 2010.
  6. Hwee Hwa Pang Jilian Zhang Kyriakos Mouratidis ,Scalable Verification for Outsourced Dynamic Databases, ACM.VLDB „09, August 2428, 2009, Lyon, France Copyright 2009.
  7. Brian Thompson, Stuart Haber, William G. Horne, Tomas Sander, Danfeng Yao, Privacy-Preserving Computation and Verification of Aggregate Queries on Outsourced Databases, HP Laboratories HPL-2009-119, published by Springer Aug-2009.

Related posts

IOT Based Water Distribution Control & Monitoring System


Handwritten Character Recognition Using Deep Learning


Performance analysis and implementation of spectrum sensing of interleaved signals in SC-FDMA using Cognitive radio


Leave a Comment