© 2017 IJSRST | Volume 3 | Issue 2 | Print ISSN: 2395-6011 | Online ISSN: 2395-602X

National Conference on Advances in Engineering and Applied Science (NCAEAS)

16th February 2017
In association with
International Journal of Scientific Research in Science and Technology

A Review on Search Scheme over Cloud Data

Akshay Shirbhate1, Aniket Waghmare2, Prof. P. H. Govardhan3

1,2
Students, Department of Computer Science and Engineering, Priyadarshini Institute of
Engineering and Technology, Nagpur, Maharashtra, India
3
Assistant Professor, Department of Computer Science and Engineering, Priyadarshini Institute of Engineering and
Technology, Nagpur, Maharashtra, India

ABSTRACT
Organizations are showing great interest in storing data on cloud. This could be the result of the unprecedented
growth of data stored in the last few years. However for the security of sensitive data, it should be encrypted before
outsourcing, which performs traditional data utilization based on plaintext keyword search. Because of the several
numbers of data owners and documents in the cloud, it is required to allow multiple keywords in the search request
and retrieve documents in the order of their relevance to these keywords. So far, several works have been studied to
achieve various search functionality, such as single keyword search, multi-keyword Boolean search, ranked search,
multi-keyword ranked search, etc. Among them, multi key word ranked search found more efficient to sort the
search results.
Keywords : Sensitive Data, Single Keyword

I. INTRODUCTION cloud platform to improve their business operations,

research, or everyday needs. With the remunerative
Currently we have a tendency to square measure in option of pay-asyou- use, general and private knowledge
associate degree information explosion era wherever square measure outsourced by several individual users
perpetually getting new hardware, software and and organizations to 3rd party CSPs. A knowledge
coaching IT skilled is turning into a nightmare for owner will source their knowledge to the cloud and
virtually each IT person. Coincidentally, we square either he will question on it outsourced data or will attest
measure witnessing associate degree enterprise IT a shopper to perform question.
design that shifted to a centralized, more powerful
computing paradigm famous as Cloud Computing, in II. LITERATURE SURVEY
which enterprise’s knowledge bases and applications
square measure stirred to the servers within the massive 2.1 Secure and privacy preserving keyword
data centres (i.e. the cloud) managed by the third-party search
cloud service providers (CSPs)a in the web. Cloud
computing has been recognized as the most momentous
Qin Liuy [1] proposed in this paper that the search
turning purpose within the development of data
provides keyword privacy, data privacy and
technology throughout the past decade. People square
measure attracted by the advantages it offers, such as linguistics secure by public key coding. CSP is
personal and versatile access, on-demand computing involved in partial decoding by reducing the
resources configuration, considerable capital communication and machine aerial in secret writing
expenditure savings, etc. Therefore, many firms, method for finish users. The user submits the
organizations, and individual users have adopted the keyword trapdoor encrypted by users private key to

NCAEAS2304 | NCAEAS | ACET | January-February-2017 [(3)2: 10-12 ]

10
 Volume 3 | Issue 2 | IJSRST/Conf/NCAEAS/ACET/2017/04

cloud server firmly and retrieve the encrypted 2.4 A Secure and Dynamic Multi keyword
documents. But the communication and machine
price for encryption and secret writing is much Ranked Search Scheme over Encrypted Cloud Data
higher. This proposed methodology [4] by Z. Xia suggest a
secure tree-based search theme over the encrypted
2.2 Secure and Efficient Ranked Keyword cloud storage, which supports multi keyword
Search hierarchal search on with dynamic operation on
document assortment obtainable at server. The
Cong Wang [2] proposed stratified search that will vector space model and term frequency (TF) ×
increase system usability by sanctioning search inverse document frequency (IDF) model area unit
result connection ranking instead of causing uniform jointly used in the development of index and
results, and further guarantees the file retrieval generation of question to supply multi keyword
accuracy. Specifically, they explore the statistical hierarchal search output. To obtain high search
live approach, i.e. relevance score, from information potency results, author construct a tree-based index
retrieval to build a secure searchable index, and structure and proposed a Greedy Depth-first Search
develop a one-to-many order-preserving mapping algorithmic program based mostly on this index
technique to properly secure those sensitive score tree. Because of this special structure of treebased
information. The resulting style is in a position to index, the proposed search themeill flexibly
facilitate economical server-side ranking while not accomplish sub linear search time and will
losing keyword privacy. But it will not perform effectively handle the deletion and insertion of
multiple keyword searches. documents. The kNN algorithm is applied to write
in code the index and question vectors, and till then
2.3 Efficient and Secure Multi-Keyword guarantee correct connection score alculation
between encrypte index and question vectors.
Search on Encrypted Cloud Data This proposed
technique by C. Orensik has defined and resolved III. TECHNIQUES AND ALGORITHMS
the drawback of effective however safe and sound
hierarchal keyword search over Encrypted cloud Some of the models, techniques and algorithms
knowledge [2]. Ranked search greatly enhances being used within the existing system are mentioned
system usability by returning the matching files in a and summarized as follows.
hierarchal order concerning to bound vital criteria
(e.g. keyword frequency) thus creating one step 3.1 Vector area Model
nearer towards wise consumption of secure
knowledge hosting services in Cloud Computing. This model is used to represent the text by a vector
These pers has defined and resolved the difficult of functions. The terms are the words and phrases. If
drawback of privacy protective and economical words are thought of as terms, every word becomes
multi keyword hierarchalsearch over encrypted associate freelance dimension in a terribly high
cloud knowledge storage (MRSE), and establish a dimension vector area. If term represents a text, it
set ofstrict privacy requirements for such a otected gets a non- zero value n the text-vector on the
cloud knowledge utilization system to become a dimension like the term. Text vectors are terribly
reality. The proposed ranking technique proves to be area and no term is allotted a negative worth.
economical to go back extraordinarily relevant
documents admire submitted search terms. But it
lacks dynamic updation and deletion of the
document from the cloud.

International Journal of Scientific Research in Science and Technology (www.ijsrst.com)

11
 Volume 3 | Issue 2 | IJSRST/Conf/NCAEAS/ACET/2017/04

3.2 Searchable secret writing rule encrypt knowledge and send it to the server. The
owner provides decryption key might be completely
An rule that consists of the polynomial time different.
randomized algorithms. They are: eyGen(s) - s is a
security parameter taken and wont to generate a key IV. CONCLUSION
pair either public or personal. PEKS (Apub, w) -
Apub may be a public key and w is a word which In this review paper, we have summarized totally
square measure wont to manufacture a searchable different reasonably looking out techniques for the
secret writing. encrypted information over cloud. A systematic study on
the privacy and data utilization problems is roofed here
for numerous looking out techniques. Some of the
Trapdoor (Apriv, w) - Apriv may be a private key
important roblems to be handled by the looking out
and w is a word that square easure wont to
technique for providing the information utilization and
manufacture a trapdoor Tw. security area unit keyword privacy, Data privacy, Fine-
grained Search, Scalability, Efficiency, Index privacy,
3.3 Term Weight Query Privacy, Result ranking, Index confidentiality,
Query confidentiality, Query Unlinkability, semantic
Term weighting is a technique that depends upon security and Trapdoor Unlink ability. The limitations for
the higher estimation of assorted chances. The main all the searching techniques mentioned during this paper
three factors play in term weight formulation is: area unit mentioned in addition. From the above survey,
Term Frequency - Words that repeat multiple times we will conclude that multi-keyword stratifiedb search
theme provides the economical, secure and dynamic
in a document. Document Frequency - Words that
search results.
appear in several documents square measure thought
of common. Document Length - hen assortment
V. REFERENCES
have documents ofvariable lengths, longer
documents influence to score higher since they
[1]. Qin Liuy, Guojun Wangyz, and Jie Wuz, ”Secure
contain additional words and more repetition. and privacy preserving keyword searching for
cloud storage services”, ELSEVIER Journal of
3.4 Cipher Text Security Network and computer Applications, March 2011
[2]. Cong Wang et al., ”Enabling Secure and Efficient
It is a method that's wont to provide security for the Ranked Keyword Search over Outsourced Cloud
encrypted knowledge. A cipher text attacker might Data”, IEEE Transactions on parallel and
simply break linguistics security by rearrangement distributed systems, vol. 23, no. 8, August 2012
the keywords and submitting the ensuing cipher text [3]. C.Wang, N. Cao, K. Ren, and W. Lou, “Privacy-
preserving multi-keyword ranked search over
for secret writing. A standard technique is employed
outsourced cloud data,”Parallel and Distributed
to interrupt this and this system is termed the cipher
System, IEEE Transction on ,vol.23,no.
text security.
8,pp.1467-1479,2012
[4]. Zhihua Xia, “A Secure and Dynamic Multi-
3.5 Non-Public Key Searchable Encoding keyword Ranked Search Scheme over Encrypted
Cloud Data”, IEEE Transactions on parallel and
A model called non-public key searchable encoding distributed systems, vol: pp no: 99 year 2015
is used to look on a personal key encrypted
knowledge. The user himself encrypts data, so as to
organize in Associate in Nursing impulsive manner.
3.6 Public Key Searchable encoding Public key
searchable cypherion is a model that permits user to

International Journal of Scientific Research in Science and Technology (www.ijsrst.com)

12