Privacy Preserving Data Mining Using Inverse Frequent ItemSet Mining Approach

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 02| Feb -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 897
Privacy Preserving Data Mining Using Inverse Frequent ItemSet Mining
Approach
Ms. Ashwini S. Chavan1, Prof. Rahul P. Mirajkar2
1Student, Department of Computer Science and Engineering, Bharati Vidyapeeth’s College of Engineering,
Kolhapur, Maharashtra, India
2Assistant Professor, Department of Computer Science and Engineering, Bharati Vidyapeeth’s College of
Engineering, Kolhapur, Maharashtra, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The paper presents architecture for protection
of information against third party attack. Individual
sensitive information is in danger with increasing
technologies of data mining. A new research data mining
topic, known as privacy-preserving data mining (PPDM),
has been tremendously studied in recent years. Privacy
preserving data mining (PPDM) aims to maintain privacy of
individual data or sensitive information without sacrificing
the utility of the data. Currently, privacy preserving data
mining (PPDM) mainly consciousness on a way to reduce the
privacy threat delivered by way of data mining operations,
even as in truth, unwanted disclosure of sensitive
information may manifest in the system of data collecting,
data publishing, and information (i.e. The data mining
effects) delivering. To this effect, paper proposes a technique
called inverse frequent itemset mining approach that will
help to protect sensitive information without loss of data.
Key words: Data mining, sensitive information,
privacy preserving data mining (PPDM), inverse
frequent itemset mining, protection.
1. INTRODUCTION
Mining of data is the process of discovering
interesting patterns and knowledge from large amounts of
data [1]. The information collected by data mining can be
very important to many applications, despite that there is
another concern on focus of the privacy threats posed by
data mining [2]. To address the privacy troubles in data
mining, Privacy preserving data mining (PPDM) has
received a top notch improvement in latest years[3][4].
The objective of PPDM is to safeguard sensitive
information against the disclosure of data by maintaining
its utility.
The PPDM Consideration is two-fold. First,
sensitive raw data e.g. individual’s ID card number, cell
phone number should not be directly used for mining.
Second, sensitive mining results whose disclosure will
cause privacy violation should be excluded.
In PPDM process four different types of users are
involved namely data provider, data collector, and data
miner and decision maker. Each one has their specific role
in the process. A data provider owns a few data from
which precious information can be extracted. Data
collector gets data from data providers so as to support
the subsequent data mining operations. In order to invent
useful knowledge which is expected by the decision
maker, the data miner applies data mining algorithms to
data obtained from data collector. A decision maker can
get the data mining results directly from the data miner, or
from some Information Transmitter. The focus of this
topic is to achieve privacy at data miner level. Data miner
will get data to mine from data collector to be able to in
not unique layout and by means of applying one-of-a-kind
data mining strategies; data miner can discover sensitive
information. So venture of data miner is to hold the
privacy of received result and pass the consequences to
decision maker that doesn't bring about any security
breach. Several studies on PPDM have been conducted [5]
[6]. But none of the modern-day proposals provide privacy
to unwanted disclosure of sensitive information. The
paper presents a system architecture that provides
privacy by the use data mining algorithms without
affecting the security of sensitive information contained in
the data.
2. LITERATURE SURVEY
Several theoretical approaches for privacy preserving
of data have been proposed in the literature.
2.1 Related Research
B. Fung, K. Wang, R. Chen, and P. S. Yu et al [6]
introduced techniques to protect the data. Data in its
original form have sensitive information about person,
and publishing such data will violate individual privacy.
Privacy-preserving data publishing (PPDP) explains
methods and tools for publishing useful information while
preserving data privacy. The author introduces different
schemes to PPDP, study the challenges in practical data
publishing, and clarify the differences and requirements
that distinguish PPDP from other related problems.
T. Mielikainen [7] introduces a well known
technique called frequent set mining to describe binary
data. However, it is an open problem how difficult it is to
make opposite the frequent set mining. The author
analyze the computational complexity of the problem of

finding a binary data set consistent with a known
collection of frequent sets and show that in many cases the
problem is computationally very difficult.
M. Malik, M. Ghazi, and R. Ali [8] introduced Privacy
preserving data mining techniques to protect individual
data or sensitive knowledge without giving up the utility
of the data. People have become used to the privacy
intrusions on their personal data and are very disinclined
to share their sensitive information. This may cause to the
accidental results of the data mining. So, the authors
introduced present current scenario of privacy preserving
data mining techniques and propose some future research
directions.
S. Matwin [9] presents a brief summary and review of
Privacy-preserving Data Mining (PPDM). The review of the
existing approaches is built along an uncertain taxonomy
of PPDM as a field. The main axes of this taxonomy specify
what type of data is to be protected, and what is the
possession of the data (centralized or distributed). Author
discusses the relationship between PPDM and preventing
unfair use of data mining techniques and round up by
discussing some of the new, arising challenges before
PPDM as a field.
K.Sathiyapriya and G. S. Sadasivam [10]
introduced various techniques to protect the data. Data in
its original form contains sensitive information. This
paper introduces methods to preserving privacy of
association rule mining and some hiding algorithms of
association rules are evaluated. Some information is
important with respect to opposing concept in
organizations and individual misuses. Nowadays in order
to discover the useful patterns in a great amount of data,
data mining tools are used. In order to protect
information, fast processing and preventing from
disclosing private data to keep privacy is presented in data
mining.
2.2 Existing System
The information spotted by data mining can be very
vital to many applications. People have shown increasing
concern about the other side of the coin, namely the
privacy threats posed by data mining. Individual’s privacy
may be disturbed due to the unauthorized access to
personal data, the undesired discovery of one’s
embarrassing information, the use of personal data for
purposes other than the one for which data has been
collected, etc. Current models and algorithms proposed for
PPDM mainly focus on how to hide this sensitive
information from certain mining ways. Besides the mining
phase, privacy issues may also arise in the phase of data
collection or data pre-processing, even in the delivery
process of the mining results.
2.3 Limitations of the Existing System
Most current studies only manage to achieve
privacy preserving in a statistical sense. Current studies
are still in the theoretical stage. By studying different
papers some observations have been drawn. Developing
practical personalized anonymization methods is in urgent
need. Besides, introducing personalized privacy into other
types of PPDP/PPDM algorithms is also required.
Therefore, it is necessary to develop new system to specify
what kind of information is required and how to present,
store, acquire and utilize the information.
3. PROPOSED SYSTEM
As of now, maintaining privacy of the data is a
difficult task. A number of methods and techniques have
been proposed for privacy preserving data mining. The
theoretical studies studied earlier defined various
techniques for protection of sensitive information. But
there yet exist a practical implementation of a system that
helps for security of sensitive information. The proposed
work has following objectives,
 To define dataset and pre-processing of it.
 To transform a dataset from original state to new
state to achieve hiding of data.
 To implement Inverse Frequent item set mining
to hide data mining results.
The data set provided generally for mining purpose is in
plain data or in relational data format. This imposes
security threat on the data. To achieve Privacy in Data
Mining, the data provided for mining purpose should not
be delivered as it is to the data miner. The Privacy
Preserving Data Publication (PPDP) is one of the
techniques that provide some data transformations and
avoid exposure of sensitive data. This transformed data is
then provided to data miner. The mining of data produces
some useful results that can be used for knowledge
extraction. But it is important to hide these results for
privacy reason.
The mining phase here uses apriori algorithm and
finds the frequent items from transformed database. To
achieve privacy for mining result some data customization
is needed and that can be provided by Inverse Frequent
Itemset mining process. This will avoid exposure of mining
results from data miner and decision maker.
The architecture consists of four modules:
1. Data preprocessing.
2. Data transformation.
3. Data mining.
4. Data evaluation.

Fig -1: System Architecture
4. PROPOSED WORK
4.1 Data preprocessing
Data preprocessing is a vital important step in data
mining process which deals with preparation and
transformation of initial dataset. It includes basic
operations as data selection, data cleaning, and data
integration. Data preprocessing describes any type of
processing performed on raw data and to prepare it for
another processing. It modifies the data into a format that
will be more easily and adequately be processed for the
purpose of the user. A standard DB-Reduction algorithm
for data preprocessing is implemented. This algorithm
highly reduces the number of attributes and records of the
data set and improves the accuracy and decreases the
running time of the data mining algorithms in the later
stage.
Algorithm makes use of two methods namely, select
method and update method. In select method, the user can
choose records from the data set. For example, suppose e a
user needs particular students records residing in zipcode
416008, then by use of this select method the user can
acquire those needed records, instead of moving through
whole data set. This method saves the time of the user and
fetches the records as needed. Next, in update method the
records selected by the user in select method can be
updated. For example, suppose a user needs to update a
particular student record who is residing in zip code
416008, he can do this by simply making a change of zip
code from 416008 to 416009. Thus, any of the record
selected by the user can be updated.
4.2 Data transformation
Data transformation converts a set of data values
from the data format of a one system into the data format
of a data system. The goal is to transform data into forms
appropriate for the data mining operations, that is, to find
important features to represent the data. The data
collected from data provider may contain sensitive
information. To hide this information some
transformation is needed. This model use privacy
preserving data publishing (PPDP) technique which
guarantees data utility even after data amendment.
Feature selection and feature transformation are basic
operations. A standard RSA (Rivest-Shamir-Adleman)
algorithm is implemented for data transformation
technique. For example if there are tables of student
information in database, like student data and course, then
the encryption can be done on single table or multiple
tables. This algorithm is used for encryption and
decryption of data. The data from pre-processing is
encrypted so as to protect it from third party. Hence,
maintaining the security of sensitive information.
4.3 Data mining
The design employs inverse data mining such as
Inverse Frequent Set Mining (IFM) to produce data that
cannot expose sensitive information. Reverse Data
Management (RDM) which is similar to Inverse Frequent
Set Mining performs operations like compute database
input or modify an existing database input, in order to
achieve a desired effect in the output. Aprori algorithm is
implemented so as to find out the count of frequent items
from the data set. For example if you have a set of 20
queries, then aprori algorithm is used to find out the count
of frequent fields (items) being fired in a set of queries. By
doing so, list of frequent items is separated from general
data set and the remaining data is encrypted. This is
termed as Inverse frequent itemset mining approach
(IFM) approach. This makes data more secure from an
intruder attack.
4.4 Data Evaluation
This is the last module. Data evaluation include
basic operations as, identifying the truly interesting
patterns which represent knowledge, and presenting
mined knowledge in easy to understand fashion. The
results produced in data mining module are evaluated
using data evaluation module. The data obtained from
data mining module is processed in this module. Not all
the patterns found by data mining algorithms are
necessarily valid. It is common for the data mining
algorithms to find patterns in the training set which are
not present in the general data set. The final result is
represented using data evaluation module.

5. SNAPSHOTS
Encryption can be single table encryption or multiple
table encryptions.
Fig -2: Single table encryption
Fig 2: shows single table encryption method. SQL
server database is used as a database at backend. Database
school is designed under which there are tables as book,
course, student, student fees as displayed in above
snapshot. User must select any one table (book) of his/her
choice for further processing. After this, fieldnames is to
be selected to create a new table so as to perform
encryption on that table. Here fieldnames branch, title and
publisher is been chosen. User must select at least two
fields to create a table. She/he can also select all the fields
to create a new table. After table creation (Encrypt_book)
the user has to select the fields (title) from new table
(Encrypt_book) on which encryption can be done. Finally
we will get data in encrypted format. User can also
transfer data as it is i.e. without encryption to the newly
created table.
Fig -3: Multiple table encryptions
Fig 3: shows multiple table encryptions. As for
single table, same process is carried out for multiple table
encryptions. In this user can select more than one table
and use it for encryption as shown in above snapshot.
Table’s course and student are selected. Fieldnames from
both tables are displayed, from which user selects sname,
rno and cname. Table Encrypt_student_course is created
consisting of fields selected by user. At last, field that is to
be encrypted is chosen and finally get message of success.
6. RESULT ANALYSIS
In this section, we calculate the time required for
encryption. The encryption is done on single table and
multiple tables.
6.1 Single table encryption
The encryption is performed on single table of the
database and the time required for it is calculated.
Chart -1: Single table
The time taken is in ms as shown in chart 1. The
time taken for encryption of 2 fields is 5ms and that for 3
fields is 7ms and so on.
6.2 Multiple tables encryption
The encryption is performed on multiple tables of
the database and the time required for it is calculated.
Chart -2: Multiple table

The time taken for multiple fields’ encryption is
more as compared to that of single table encryption.
7. CONCLUSIONS
We have made survey of different literatures and
studied different techniques for maintaining privacy of the
data. This paper proposes development of a system that
will help to protect the sensitive information. Various
algorithms are being used by the modules to protect a
system against a theft. The data is protected in four steps
as discussed in section IV. Inverse Frequent Itemset
mining approach is used to achieve privacy. Due to this,
security of data is maintained without losing the original
essential information.
8. FUTURE WORK
As future work, we can implement this same system
by using other data mining technologies other than apriori
algorithm like FP-Growth algorithm for finding frequent
items in large database. Also we can use an improved
aprori algorithm. For further research we can go through
additional reference paper which can helpful to look on
advanced research like. Multi-Sorted Inverse Frequent
Itemsets Mining: On-Going Research. It is needed that
design the database more detail as per needed by
emerging big data applications e.g., social network
analytics. We can increase the accuracy of algorithm in
spite of large database.
ACKNOWLEDGEMENT
There have been many contributors for this to take shape
and authors are thankful to each of them.
REFERENCES
[1] Lei xu, Chunxio Jiang, Jian Wang, Jian Yuan, Yong Ren.
Information Security in Big Data: Privacy and Data
Mining, DOI 10.1109/ACCESS.2014.2362522,IEEE
Access.
[2] J. Han, M. Kamber, and J. Pei, Data mining: concepts
and techniques. Morgan kaufmann, 2006.
[3] L. Brankovic and V. Estivill-Castro, “Privacy issues in
knowledge discovery and data mining,” in Australian
institute of computer ethicsconference, 1999, pp. 89–
99.
[4] R. Agrawal and R. Srikant, “Privacy-preserving data
mining,” SIGMOD Rec., vol. 29, no. 2, pp. 439–450,
2000.
[5] Y. Lindell and B. Pinkas, “Privacy preserving data
mining,” in Advances Cryptology CRYPTO 2000.
Springer, 2000, pp. 36–54.
[6] B. Fung, K. Wang, R. Chen, and P. S. Yu, “Privacy-
preserving data publishing: A survey of recent
developments,” ACM Computing Surveys (CSUR), vol.
42, no. 4, p. 14, 2010
[7] T. Mielikainen, “On inverse frequent set mining,” in
Workshop on Privacy Preserving Data Mining, 2003,
pp. 18–23.
[8] M. Malik, M. Ghazi, and R. Ali, “Privacy preserving data
mining techniques: Current scenario and future
prospects,” in Computer and Communication
Technology (ICCCT), 2012 Third International
Conference on, 2012, pp. 26–32.
[9] S.Matwin,“Privacy-preserving data mining techniques:
Survey and challenges,” in Discrimination and Privacy
in the Information Society. Springer, 2013, pp. 209–
221...
[10] K. Sathiyapriya and G. S. Sadasivam, “A survey on
privacy preserving association rule mining.”
International Journal of Data Mining &Knowledge
Management Process, vol. 3, no. 2, 2013.
[11] Libing Wu, kui Gong, Yanxiang He, Xiaohua Ge Jainqun
Cui, ‘A Study of Improving Apriori Algorithm’ 2010.
[12] Xiaohua Hu, ‘DB-Reduction: A data preprocessing
algorithm for data mining applications’,2003
[13] A. Meliou, W. Gatterbauer, and D. Suciu, “Reverse
data management,” Proceedings of the VLDB
Endowment, vol. 4, no. 12, 2011.
[14] Domenico Saccà, Edoardo Serra, and Antonio
Piccolo."Multi-Sorted Inverse Frequent Itemsets
Mining: On-Going Research" AMW 2016 - Proceedings
of the 10th Alberto Mendelzon International Workshop
on Foundations of Data Management (2016)

Privacy Preserving Data Mining Using Inverse Frequent ItemSet Mining Approach

More Related Content

What's hot (19)

Similar to Privacy Preserving Data Mining Using Inverse Frequent ItemSet Mining Approach (20)

More from IRJET Journal (20)

Recently uploaded (20)

Privacy Preserving Data Mining Using Inverse Frequent ItemSet Mining Approach