Big data privacy and inconsistency issues

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________
Volume: 03 Special Issue: 07 |May-2014, Available @ http://www.ijret.org 812
BIG DATA: PRIVACY AND INCONSISTENCY ISSUES
ArulMurugan R1
, Anguraj S2
, Boopathi R3
1
Student, Information Technology, K.S.R College of Engineering, Tamilnadu, India
2
Assistent Professor, Information Technology, K.S.R College of Engineering, Tamilnadu, India
3
Student, Information Technology, K.S.R College of Engineering, Tamilnadu, India
Abstract
In the current Big Data world we are literally drowning in data that has been generated every day. Big data refers to data
volumes in the range of exabytes (1018
) and beyond. Such volumes exceed the capacity of current on-line storage systems and
processing systems. Data, information, and knowledge are being created and collected at a rate that is rapidly approaching the
exabyte/year range. But, its creation and aggregation are accelerating and will approach the zettabyte/year range within a few
years. Such massive volumes of data generated are new to storage industry. Many concerns have been raised to formulate the
data issues. Big Data will be the only key to overcome all these data issues. In this paper we are going to provide a closer look
towards various privacy issues that has been haunting various Big Data security later we provide an analysis about the various
inconsistencies that has an impact on Big Data analytics.
Keywords: Big Data, privacy in Big Data, Big Data security, Inconsistencies in Big Data, security issues in Bid Data.
--------------------------------------------------------------------***----------------------------------------------------------------------
1. INTRODUCTION
The Big Data has become a catch phrase in the data field
ever since the traditional databases have failed to handle Big
Data. Even though Big Data is in its early days, strive for
Big Data has already begun as we are getting ready to
process petabytes of data. Meanwhile Big Data analytics is
also reaching a different dimension as it is said to be the
next important phenomena in the data industry.
The term Big Data refers to massive collection of structured
and unstructured data that are not suitable for processing in
a traditional database. Due to which Big Data analytics has
been given affinitive importance. The Big Data is not just a
large volume of data; it is actually a data treasure with
valuable insights that are hidden inside. These insights are
the current need for different purpose starting from decision
making to scientific research. Hence the Big Data is the only
solution to address all their concerns as the current growth
of data industry has grown to an unexpected level in recent
years.
2. CHARACTERISTICS OF BIG DATA
The Granter’s Dough Laney [1] describes that Big Data has
three major characteristics namely volume, variety and
velocity. Another important characteristic that has been
under lot of speculation is the veracity. The characteristics
of Big Data are summarized in table 1.
Table -1: Characteristics of Big Data
Characteristics Description
Volume The volume refers to the amount of
data. It’s been increasing
exponentially. They are generally
data at rest.
Velocity Velocity is the speed in which the
data has been generated and
processed. They are generally data
in motion.
Variety Variety refers to the various data
formats and structures like text,
numerical data, multimedia data
etc.,
3. ORIGIN OF BIG DATA
The Big Data is generated from the various sources like
social media and networks, many scientific instruments and
business process which generate data continuous data in
different forms. Moreover temperature devices, mobile
devices, sensors track data with respect to time.
The logs such as transmission logs, server logs, and event
logs are sources of Big Data that are currently employed in
Big Data analytics.
4. EMPLOYING BIG DAT. FIELDS A
The reason behind providing more emphasis on Big Data is
that they have become an integrated factor for future
enhancement in many domains. The various sectors such as
life and physical sciences, healthcare, transportation,
banking, finance, government [5] has started employing Big
Data to extract the necessary insights.
The business areas such as innovation in business,
education, decision making must consider Big Data
analytics to stay among the competitors. The Geographical
Information Systems are widely used to identify the crime
partners [6] and the public health partner [7] employs Big
Data analytics.

_______________________________________________________________________________________
The data that has been collected from various formats from
various sources are cleansed, analyzed, summarized and
finally visualized before it can employed for Big Data
analytics to obtain insights about the particular domain.
5. ISSUES IN BIG DATA:
The Big Data comprises of several issues such as storage
issues [1], management issues, privacy issues and
inconsistency issues.
Among these issues the privacy and inconsistency issues are
the major rising up issues which are going to address.
5.1 Privacy Issues
Nowadays we share our personal information such as our
likes, dislikes and our whereabouts more frequently over the
internet. These footprints may lead to various privacy issues
like location disclosure, personal information disclosure.
5.1.1 Need for Privacy
Due to invariable increase in the internet usage the privacy
remains an unsolved mystery still now. These privacy
concerns generally occur due to illiteracy in privacy.
5.1.2 Challenges of Privacy
1) Individual privacy: When the information about a
particular person is merged together with a very large
datasets, there exists a fair possibility of identifying the
various new hidden insights about the particular individual
which he/she is unaware of [3].
2) Geo Information Privacy: While dealing with the GIS
information the location disclosure through various sources
can become a privacy threat.
3) Business Privacy: The various data and the logs such as
transaction logs, server logs can reveal confidential
information that would provide an opportunity to misuse the
information. For example if company A and B are business
competitors then both A and B must not reveal any data
between each other as this might result in loss of business
privacy.
4) Log Storage Privacy: The corporate companies usually
have to maintain huge volume of logs for future references.
If these logs accessed in improper manner, it could result in
access to denied systems.
5) Under Privileged Privacy: By analyzing the large
datasets it is easy to identify the under privileged ones. The
current social media and other resources make this much
easier. Hence under privileged privacy can be at stake.
1) Individual Attacks: The various forms of the individual
attacks are discussed below.
The major privacy issue is getting the acknowledgement
from the user whose information is taken for analytics. Here
the problem occurs when the user initially believes that
his/her personal information is secure but later if he/she
feels that no longer privacy, then various privacy laws and
data collector must guarantee the information is removed.
Next, issues that if several datasets when merged together
can hold some unique information about the individual
(user). The more focused attack is the identification of the
single entry from the dataset in such a way that the
particular individual information is obtained with higher
confidence.
In the targeted attack one or more details of a given
individual can be identified. This attack is said to have
higher impact on privacy.
2) Industry Attacks: The most common industrial attack in
privacy is creating the confusion and distraction in the
information which in turn makes the available datasets to
less useful for analyzing the business strategies.
The correlated information in the datasets may sometimes
be able to reveal the self-calculated information.
The data sets containing confusing and distracting content
can lead to improper insights that result in decreased
decision making capability. For example if company A and
B are competitors that follow same datasets. Suppose if A
passes the datasets which contains confusing and distracting
information to B. Then obviously company B’s decision
making capability will automatically decrease.
5.2 Inconsistency Issues
When several large datasets are combined together which
has a different origin the inconsistencies are bound to occur.
Moreover the real world is prone to have inconsistent data.
Generally the inconsistency in data occurs due to conflicting
information. If these inconsistencies are not handled
properly then the data sets would not have any integrity.
The inconsistencies can be a major problem in various fields
such as heuristics, problem solving, research analysis and
business prediction analysis.
In order to get a clear picture about where the
inconsistencies creep in we must first analyze the level of
the contents in big data.
5.2.1 Various Levels of Big Data
The following are the three broad levels of big data.
1) Data: The data is the preliminary level in big data. They
are basic and less structured form of the content in big data.
Generally the data are text, graphical contents, video, audio,
symbols and numeric.
For instance consider the data “1986” it may refer to year or
registration id or amount or anything that can be numeric.
Due to this character the inconsistencies are hard to identify.

_______________________________________________________________________________________
The inconsistencies at this level are very hard to identify.
The only remedy is to make sure that data is collected by
considering the integrity factor.
2) Information: The information is partially structured
hence it is to an extent easier to identify the inconsistency.
The information usually provides an extra value to the data.
Eg1: “Year: 1986” identifies the data to be a year. Hence
year can only be numeric due to which inconsistencies can
be eliminated to an extent.
3) Knowledge: They are well structured and more enhanced
representation among the contents in the data sets. Mostly
the inconsistencies are eliminated at this level but at rare
scenario the inconsistencies may occur. Hence care must be
taken while converting the content from data to knowledge.
5.2.2 Types of Inconsistencies
In the process of converting the data into knowledge the
inconsistencies must be eliminated to perform an effective
analysis. Among the various inconsistencies that are
available the most common inconsistencies are discussed
below.
1) Text Inconsistencies: The Text inconsistencies usually
originate from various sources like emails, blogs, social
media etc. These inconsistencies can highly alter the
integrity of the data. Whenever two texts are referring to the
same event or entity, then they are said to be co-reference.
The co-reference is a mandatory condition for text
inconsistencies. Various types of inconsistencies are
summarized in table 2.
Table -2: List of text inconsistencies [16]
Type Example
Synonyms
Bob is a brilliant student.
Bob is a bright student.
Antonyms
Bob is a bright student.
Bob is dull student.
Disagreeing
Bob has $100.
Bob has $1000.
Contradictory
Beta was developed in 2014.
Beta was deployed in 2013.
Complementary
Bob will graduate this year.
Bob will not graduate this year.
The other texts in consistencies are inheritance, asymmetric,
anti inverse, mismatching, etc.
2) Functional Dependency Inconsistency: The functional
inconsistency may develop if the condition for an operation
that requires the integrity constraints is not satisfied. The
functional dependencies may arise in the following three
conditions. The table 3 describes the functional
dependencies.
Table -3: Functional Dependency Inconsistencies [13], [17]
Single FD
Violation of single
functional dependency
Multiple FD
Violation of multiple
Conditional FD
Violation of conditional
3) Temporal Inconsistencies: Whenever the datasets
contains temporal attribute then the temporal inconsistencies
may occur. Since the temporal data deals with reference to
time, the conflict can occur easily. The time interval
relationship between conflicting data items can result in
partial temporal inconsistency (or) complete inconsistency.
3.1 Partial Inconsistencies: The time intervals of two
inconsistent events are partially overlapping. In a
temperature time series data, a temperature recording of -4ºc
in April in South India is inconsistent with the context.
3.2 Complete Inconsistencies: The time intervals of two
inconsistent events coincide or satisfy containment. In ECG
a prolonged period of low value output in ECG is
inconsistent with the normal heart rhythm pattern [11].
4) Spatial Inconsistencies: The spatial data consist of
geometrical properties and have various spatial relations.
The violation in the spatial constraints results in the
inconsistencies. The spatial inconsistency may occur in
reference to spatial relations such as topological, directional
and distance [14].
5.2.3 Handling Inconsistencies:
As mentioned in earlier, the inconsistencies are difficult to
handle in spite of the form of the content. The datasets must
be cleansed before it is handled so as to eliminate maximum
possibility of inconsistency occurrence.
The various artificial visualization tools, machine learning
methodologies, pattern analysis should be performed to
eradicate inconsistence in data.
6. CONCLUSIONS
In this paper we have stated various privacy and
inconsistency issues that continue to haunt Big Data
analytics. The privacy issues should be address by strictly
enforcing the privacy laws and by also creating a secure
system while handling user data.

_______________________________________________________________________________________
Similarly the inconsistencies must be eliminated as early as
possible to improve data quality there by providing efficient
data analytics.
We have planned to provide solutions for various privacy
and inconsistency issues as occur future work.
REFERENCES
[1]. Stephen Kaisler, Frank Armour, J. Albero Espinosa and
William Money, “Big Data: Issues and challenges moving
forward”, 2013, 46th Hawaii International conference on
system services, pp.995- 1004.
[2]. Du Zhang, “Inconsistencies in big data”, Proc. 12th
IEEE International conference on cognitive informatics &
cognitive computing [ICCI*CC’13], pp.61-67.
[3]. Meiko Jenson, “Challenges of Privacy Protection in Big
data Analytics”, Big Data (Big Data Congress), 2013 IEEE
International Congress on IEEE, pp. 235-239.
[4]. Linna Li, Michael Good Child, “Is Privacy still an issue
in the era of big data - Location disclosure in spatial foot
prints”
[5]. J. Manyika, M.Chui, B. Brown, J. Bughin, R. Dobbs,
C.Roxburgh, and A. H. Byers, Big Data: the next frontier for
innovation, competition, and productivity, McKinsey Global
Institute, June 2011.
[6]. Cromley, E. K., &McLafferty, S. L. (2011), GIS and
public health, Guilford Press.
[7]. Chainey. S, & Ratcliffe. J (2005), GIS and crime
mapping Wiley.
[8]. Du Zhang and M. Lu, Inconsistency-induced learning
for perpetual learners, International Journal of Software
Science and Computational Intelligence, Vol.3, No.4, 2011,
pp.33-51.
[9]. D. Zhang, i2Learning: perpetual learning through bias
shifting, in Proc. of the 24th International Conference on
Software Engineering and Knowledge Engineering, July
2012, pp. 249-255.
[10]. D. Zhang and M. Lu, Learning through Overcoming
Inheritance Inconsistencies, in Proc. of the 13th IEEE
International Conference on Information Reuse and
Integration, August 2012, pp.201-206.
[11]. V. Chandola, A. Banerjee, and V. Kumar, Anomaly
detection: a survey, ACM Computing Surveys 41 (3) 2009,
pp.1-58.
[12]. D. Zhang, On Temporal Properties of Knowledge Base
Inconsistency, Springer Transactions on Computational
Science V, LNCS 5540, 2009, pp.20-37.
[13]. M. V. Martinez, A. Pugliese, G. I. Simari, V. S.
Subrahmanian, and H. Prade, How dirty is your relational
database? -An axiomatic approach, in Proc. 9th
EuropeanConference on Symbolic and Quantitative
Approaches toReasoning with Uncertainty, Hammamet,
Tunisia, LNAI 4724, 2007, pp.103-114.
[14]. A. Rodriguez, Inconsistency issues in spatial
databases, in L. Bertossi et al (eds.) Inconsistency
Tolerance, LNCS 3300, Springer-Verlag, 2004, pp.237-269.
[15]. Stonebraker, M. and J. Hong. 2012. “Researchers' Big
Data Crisis; Understanding Design and Functionality”,
Communications of the ACM, 55(2):10-11
[16]. M-C de Marneffe, A. N. Rafferty and C. D. Manning,
Finding Contradictions in Text, Proc. of 46th Annual
Meeting of the Association for Computational Linguistics:
Human Language Technologies, 2008, pp.1039-1047.
[17]. W. Fan, F. Geerts, X. Jia, and A. Kementsietsidis,
Conditional functional dependencies for capturing data
inconsistencies, ACM Transactions on Database Systems,
Vol. 33, Issue 2, June 2008.
[18]. Matthew Smith, Christian Szongott, Benjamin Henne,
Gabriele von Voigt, “Big Data Privacy Issues in Public
Social Media”, IEEE, 6th
International Conference on Digital Ecosystems
Technologies (DEST), 18-20 June 2012.

Big data privacy and inconsistency issues

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Big data privacy and inconsistency issues (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Big data privacy and inconsistency issues