IRJET- Efficient Privacy-Preserving using Novel Based Secure Protocol in SVM

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 03 | Mar-2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 3715
Efficient Privacy-Preserving Using Novel Based Secure Protocol in SVM
Hamsaveni N1, Shere banu M2, Revanth S3, Selva Prasad J4, Sashi Rekha K5
1,2,3,4 Student, Computer science and engineering, Dr. N.G.P Institute of Technology, Coimbatore, India
5 Assistant Professor, Computer science and engineering, Dr. N.G.P Institute of Technology, Coimbatore, India
---------------------------------------------------------------------***------------------------------------------------------------------
Abstract - Data Mining is the practice that categorizes large
datasets to extract the meaningful information. Data are
analyzed and segmented in the database that is helpful in
identifying the previously hidden patterns of raw data.
Support Vector Machine is used for classifying the data in the
database Server. Users data will be classified and theresultsof
classification be sent to the user which suffers from data
leakage. Given the importance of maintaining security of
original data, we proposed a new novel based framework
protocol proclaimed as Light Weight Multiparty Random
Masking and Polynomial Aggregation Protocol. In our work,
the SVM classification result serves more Accuracy, Efficiency
and also maintains the confidentiality of user’s data.
Key Words: data mining, classification, cloud computing,
support vector machine, privacy preserving
1. INTRODUCTION
In today’s era where the technologies are emerging widely
in various Phases and in different fields and forms, an extra
pinch is added by data mining. The technological
development has marked a point in which a lead taken by
data mining to provide a perk to artificial intelligence and
machine learning immersing along with it. Also, owing to
massive increment of the datasets, retrieving of such data
and maintaining is formidable. So, the most Influential and
heft technique used for extracting and culling the
information from the huge database is the data mining. Data
mining is an epitome which involves large scale of datasets
and associates to cluster, classify, and detect the anomalies
and henceforth summarizing the data. As classifyingthedata
is essential in data mining there is a major issue in
maintaining them, this directs to the introduction ofsupport
vector machine (SVM).
Considering the significance of mining and decision
making, the organization and institution requires support
vector machine to provide better classificationresultoftheir
data. Our System comprises of three entities that include a
cloud Computing, server and data provider.
Privacy is the key aspect which encloses sensitiveand
primitive form of information about the users. Managing of
the data in the big organizations is done mostly through the
cloud computing technology. All though, cloud computing is
very popular for the purpose of the storage; the loop hole
here is security. To prevent the Jeopardizing of data, the
encryption is done using novel based framework such as
Light weight multiparty random masking and Polynomial
Aggregation protocol.
2. EXISTING WORK
In existing work, a protocol is used to securelydeterminethe
classification result. In addition, existing schemeuseshybrid
approach utilizing the combination of homomorphic
encryption system and garbled circuit protocol. The existing
scheme can also be severed as an efficient basic building in
any other application that needs to figure out the sign of a
number in encrypted domain. TheSecureSignProtocol(SSP)
used in the existing system which can securely and correctly
work out the sign of the user. SSP discloses nothing useful
about the privacy of any participant, and is probably secure
only when it deals with small length of the users keys. With
the growth of l which is a positive integer used in the user’s
key, SSP can save more computation time. If the value of l is
high, the SSP is consuming more time.Thus,ourcomputation
time increases, as the key is longer. Nevertheless, the value
of l has little influence on computation time, since the size of
the garbled circuit protocol does not vary with l. For a fixed
key, the computation time of SSP almost remains the same,
as l varies. Therefore, in the existing work, SSP is not having
efficiency to securely determine the sign in encrypted
domain1.

3. PROPOSED WORK
In proposed work, a novel based approach is used for the
data security. A novel based protocol is the combination of
two protocols called Light weight Multiparty Random
Masking and Polynomial Aggregation Privacy-Preserving
protocol. This two protocols works separately to enhance
the data security. In proposed scheme, the user’s data are
encrypted using the Light weight Multiparty Random
Masking protocol. Then, data leakage is prevented using the
Polynomial Aggregation Privacy-Preserving protocol. This
protocol increases the efficiency and enhances the security
of data.
3.1 DATA MINING
Data analysis plays major role in data mining. Refinement of
results through mining is due to analysis of data in database.
Data mining is one of the inter-disciplinary fields which are
made up of combination of statistics, technological database
and artificial intelligence. The presentable work of mining
for the custom data is used for predicting the future and
explaining the past by means of analyzing the data in the
huge database. Usage of database by business organizations
are widely increasing due to increased amount of the data
over years and the Users. This leads to the data mining tools
and similarly, application ratio of data mining is also
estimated to be very high at most of the time. Hence, data
mining is used for extracting the most knowledgeablewhich
is also very valuable from the datasets.
The implementation of mining gives the refinement of
data from the huge amount of datasets. Data stored in the
database is not only involving with single user also
comprises of combination of business organizations and
institutions. This does not give assurance to the security2 of
data. Sometimes the data reflects the thorough information
about any organizations or users private data.
3.2 CLOUD COMPUTING
The cloud computing is used for maintaining the
intercommunication which includesthe user and the server.
In daily life, internet isan important tool eitherpersonallyor
professionally. Due to movement, presence and cheap, the
cloud computing is highly known. The main purpose is to
extract structure information fromtheunstructuredorsemi-
structured data resources3. Cloud security design is active
only when the right gateway implementations are in
location. The main intension of this controlling are to
minimise the hacking on the cloud computing. In proposed
work, the data which needs to be classified for the user send
the data to the server. After the data has been classified by
using SVM classifier4, results of classification gets stored in
the cloud.
3.3 SUPPORT VECTOR MACHINE
Support vector machine is a supervised machine learning
algorithm. SVM fulfillsthe needsof both classificationaswell
as regression. A SVM is normally used to segregate the data
and classify them into datasets .SVM6 classifies the datasets
and eventually labels them according to the user’s
requirement. In this particular paper the data is sent for a
text classification in which the data is segregated and
henceforth labeled. The labeled data is taken and using the
algorithm it is encrypted. In previous work also SVM was
used for purpose of classification. However here SVMplaysa
major role in the case of segregation.SVM is therefore used
not only in technological stuffs it is also used in other fields
as well.
4. MECHANISM
In proposed, a new novel framework based on lightweight
multiparty random masking and polynomial privacy-
preserving is proposed for improving the efficiency and
privacy of protocol. Furthermore, the proposed framework
can efficiently protect the data privacy as well as ensure
confidentiality. Specifically, for a data query from a
registered user, the response isdirectlyperformedoncipher
text at the service provider without decryption, and the
prediction result can also only be decrypted by the
registered user, meanwhile the query result is consistent
with that of un-privacy preserving scheme. The Detailed
security analysisdemonstratesourproposedprotocolshows
its security strength and privacy-preserving ability, and
extensive experiments are conducted to demonstrate its
efficiency.
In this work, connection is established among the client,
cloud and server. Then, either the user who wantsto classify
his/her data or any business/institutions that needs
classified data send their unprocessed data to SVM for
classification. Initially, the users select the file which
containsthe primitiveform data and thenuploaded.Thetask
of file selection and transferring has been takes place in the
user’sside. These unprocessed data flowsfromuser’ssideto
the server’s side for classification. During the transfer stage,
there is a chance of hacking the client’s private data by
intruders. So, our proposed Novel Based Protocol helps in
preventing the client’s data. Light Weight Multiparty
Random Masking protocol is used for encrypting the data
and the encrypted data been transferred from user’s side to
the server’sside for classification. SupportVectorMachinein
the Server side classifies the encrypted data. The
classification results then gets stored in the cloud for the
retrieval process of user’s need. The important task is to
prevent the security of data. Here, Polynomial Aggregation
Protocol prevents the data leakage and maintaining the
privacy data in a secure way.
5. RESULT AND ANALYSIS
The performance is tested with two real time data set in
terms of accuracy and time. The result shows the proposed
privacy preserving method perform efficiently better than
the conventional methods.
Accuracy: The accuracy gives true results in proportion
(both true positives and true negatives) among the total

number of cases examined. Accuracy can be calculated from
formula given as follows:
Accuracy
Time: The indefinite continued progress of existence and
time taken for classify the encrypted data.
Time
Precision- Precision value is evaluated according to the
feature classification at true positive prediction; false
positive.It is expressed as follows:
Recall- Recall value is evaluated according to the feature
classification at true positive prediction, false negative. It is
given as,
6. CONCLUSION
In the following paper, soundness and leakage wasthe main
issue and apart from that encryption was carried out using
secure sign protocol which mainly focused on the matching
of the signs to classify and then label the data. In the
proposed model we use light weight multiparty random
masking protocol which does not require any matching of
the key, also classification is easier .Therefore, comparing
with the earlier paper we achieve more accuracy ,efficiency
and assuring no leakage of the data possible by any means.
REFERENCE
[1] Xingxin Li, Youwen Zhu, Jian Wang, Zhe Liu, Yining Liu,
Mingwu Zhang On the Soundness and Security of
Privacy-Preserving SVM for Outsourcing Data
Classification IEEE Transactions on Dependable and
Secure Computing ( Volume: PP, Issue: 99 ) March 2017
[2] Yu, Y., Li, Y., Yang, B., Susilo, W., Yang, G., & Bai, J. (2017).
Attribute-Based Cloud Data IntegrityAuditingforSecure
Outsourced Storage. IEEE Transactions on Emerging
Topics in Computing.
[3] Liu, X., Choo, R., Deng, R., Lu, R., & Weng, J. (2016).
Efficient and privacy-preserving outsourced calculation
of rational numbers. IEEE Transactions on Dependable
and Secure Computing.
[4] Malina, L., & Hajny, J. (2013, July). Efficient security
solution for privacy-preserving cloud services.
In Telecommunications and Signal Processing (TSP),
2013 36th International Conference on (pp. 23-27).
IEEE.
[5] Omer, M. Z., Gao, H., & Sayed, F. (2016, November).
Privacy Preserving in Distributed SVM Data Mining on
Vertical Partitioned Data. In Soft Computing & Machine
Intelligence (ISCMI), 2016 3rd InternationalConference
on (pp. 84-89). IEEE.
[6] Rahulamathavan, Y., Phan, R. C. W., Veluru, S., Cumanan,
K., & Rajarajan, M. (2014). Privacy-preserving multi-
class support vector machine for outsourcing the data
classification in cloud. IEEE TransactionsonDependable
and Secure Computing, 11(5), 467-479.
[7] J. Clerk Maxwell, A Treatise on Electricity and
Magnetism, 3rd ed., vol. 2. Oxford: Clarendon, 1892,
pp.68-73.

[8] I.S. Jacobs and C.P. Bean, “Fine particles, thin films and
exchange anisotropy,” in Magnetism, vol. III, G.T. Rado
and H. Suhl, Eds. New York: Academic, 1963, pp. 271-
350.
[9] K. Elissa, “Title of paper if known,” unpublished.
[10] R. Nicole, “Title of paper with only first word
capitalized,” J. Name Stand. Abbrev., in press.
[11] Y. Yorozu, M. Hirano, K. Oka, and Y. Tagawa, “Electron
spectroscopy studies on magneto-optical media and
plastic substrate interface,” IEEE Transl. J. Magn. Japan,
vol. 2, pp. 740-741, August 1987 [Digests 9th Annual
Conf. Magnetics Japan, p. 301, 1982].
[12] M. Young, The Technical Writer’s Handbook. Mill
Valley, CA: University Science, 1989

IRJET- Efficient Privacy-Preserving using Novel Based Secure Protocol in SVM

More Related Content

What's hot (20)

Similar to IRJET- Efficient Privacy-Preserving using Novel Based Secure Protocol in SVM (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- Efficient Privacy-Preserving using Novel Based Secure Protocol in SVM