Role of Big Data Analytics In Banking

Tushar Gupta1 Naman Gupta2 Ankit Agrawal3 Aksh Agrawal4 and Kartik Kansal5
Department of Information Technology
ASET, Amity University
NOIDA, UP, India
1
[email protected] [email protected]
3
[email protected] [email protected] [email protected]

Abstract² Big data analytics is one of the IT industries trending Analytics is basically the term which we use to define the
and popular field. As the population is increasing at a constant process of taking in the data and after the execution apply
rate, it will also in turn lead to increase in the volume of the data some mathematical or statistical techniques or even some tools
which we refer it as big data. Nowadays, various organizations or to determine the useful information from the information
industries (like banking, e-commerce, insurance, etc.) are stored.
working on analyzing this big data by dividing the raw data in
Big data analytics is derived from big data which can be
several fractions so that these sectors can facilitate their
customers by analyzing each fraction efficiently. Analysis of big explained as the process of examining large volumes of data
data help these industries in various fields like, money through which industries as well as various organizations (like
movements, threats, disasters, etc. banks, e-commerce, insurance, etc.) to know different
Nowadays, online banking is one of the general service which was LQGLYLGXDO¶V VSHQGLQJ SDWWHUQ FRUUHODWLRQV HWF ZKLFK KHOSV
provided to all the customer. As the e-transactions are increasing these industries to find out new opportunities which ultimately
due to online banking facilities risks of fraud is also increased, so leads to smarter business moves, effective decision making,
big data analytics also helpful in determining the detection of higher profits and happier customers.
fraud. Every banking industry generates crucial data from the What is the need of the big data? What do customer need from
customer which needs to be stored and analyzed effectively using
their banks?
big data analytic methods to get the necessary insights for
banking organizations. Nowadays banking industry start Traditional tools are not capable of providing such
working on analyzing big data to attain goals of marketing. information. So, big data is invented to provide the way of
Hence, use of BDA in India would help banks in generating the handling this data so that industries can increase their growth
actions which would help in making the future decisions more rate. Big data also helps in providing the path to help us in
effective and stay at peak in business and competition. revolutionary transformations in several fields like inventions,
Keywords²Big data, raw data, online banking, BDA marketing statistical status, etc. we have also concluded that
with the help of big data analytics several organizations have
I. INTRODUCTION tremendously benefited from analyzing the big data to achieve
good raw data from it. Big data analytics plays a vital role in
%HIRUH XQGHUVWDQGLQJ ZKDW LV ³ELJ GDWD DQDO\WLFV´" it is bank industries as the amount of data of bankLQJ FXVWRPHU¶V
necessary to know about what is big data and its sources as increasing rapidly, it will affect the quality of service offered.
well. Basically big data is defined as the collection of It helps by making the task simpler as whenever the customer
information from various resources which is large in volume name or account number is entered in the system, it examines
as well as complex in nature. Also it can be explained as the all the information available and provides only the required
collection of data which is big in volume and it is not so easy data. This allow the banks to streamline their process of
to store and process. working and save both time and cost.
Sources of big data: The main challenge which the banking industry has to face the
risk at accounts as soon as possible. Also big data helps in
analyzing the account through which banks are able to know
repaying capacity of the borrower. With the help of big data
EDQNV FDQ H[DPLQH WKH LQIRUPDWLRQ WR WUDFN WKH FXVWRPHU¶V
behavior in real time, which would help the industry that what
amount of requirements are needed at any instant of time.

Fig1: Various sources of vast amount of data[11]

c
978-1-7281-5529-6/19/$31.00 2019 IEEE 222
 II. LITERATURE REVIEW A. Characteristics of Big Data
In the review we defined big data as combination of Big data was ideally a new concept, it has very vast number of
information whose size is very large in volume or in content. definitions derived from researchers, industries, and individual
Raw data software tools are used for gathering, storing, consumers as well. In the year 2001, industry expert explained
examining and analyzing. In the report ³5ROH Big Data WKH WHUP ELJ GDWD LQ UHJDUGV ZLWK WKH WKUHH 9¶V ZKLFK DUH
Analysis LQ %DQNLQJ´ ZH VWDWH that source must be classified Volume, Velocity, and variety. SAS (statistical Analysis
into big data in three Vs: volume, variety and rate. Most system) has given two new category which are Variability and
people thinks that big data basically has volume, but many the complexibility. Later on, Oracle defined big data in regard
VWDWHV WKDW LW FOHDU WKDW WKH WZR 9¶V DUH MXVW DV LPSRUWDQW %LJ with the help of four V's i.e. Volume, Velocity, Vast nature
Data doesQ¶WRQO\ELJEXWLWLVDOVRYHU\YDVWLQWKHQDWXUH,WLV and Value .Later on, Oguntimilehin.A, explained the concept
available in many formats and could be well presented in of big data in terms of five V's Volume, Velocity, Variety,
structured or unstructured category. Rate refers to the speed Variability, Value and a Complexity. In the year 2014, Data
through which it was generated. He suggested that the other Science Central, Kirk Bornhas briefed big data with the help
ways to calculate the volume of information would be the RI  9¶V LH 9ROXPH 9DULHW\ 9HORFLW\ 9HUacity, Validity,
content amount of files, records, payments, etc. Value, Variability, Venue, Vocabulary, Vagueness[5]. These
In the Glossary they explained big data as large-volume, high- traits has provide a path in research medium to the researcher
rate and vast variety. Information sources which needs less as well as practitioners in order to facilitate the information
expensive, innovative forms of raw data processing for effectively. The complete research on big data reveals around
efficient insights and affective decision making process. For such traits in order to affectively analyze and use big data. But
the conclusion at "Discovering Big Data: A Practical professor some areas are still left which needs to be defined in order to
to make change in the working of various organizations, big achieve healthy insights in the area.
data department of the TechAmerica for foundation in big data
B. Research Work in Big Data Traits
is a homemade name which describes it as high volumes of
increasing rate, which was complex and has vast variety of 9DULRXVLQGXVWULHVLQWRGD\¶VPRGHUQZRUOGDUH handling large
data that needs latest methods and technologies for gathering, quantities of the data. This traits generates several issues. But
preventing, spreading, managing and analyzing raw various solutions are also now available which helps us to
information. Use of big data analytics at various organizations recover from this difficult situation. Due to this immense
like banking industry proves to be very useful which was problem it generates greater need for industries like banking
easily understand as follows: industry to get brief of big data. There are more traits which
12% banks are analyzing big data.25% banks are increasing help in greater understanding of the term big data which are
their growth rate by processing it.38% banks are expanding defined as
various paths. EDQNV DUH GRLQJ PDQ\ SUDFWLFDO¶V DERXW Verbosity: Big data comes from very vast sources which are
how we are able to use it in the vast diversity. present in different forms namely structured or unstructured
Resources for banks: information. So, we explain verbosity as the various formats in
Credit Card History: To track the widely used gateways. which the raw information is available from diverse resources.
Payments: analyzing of loyal customers. Voluntariness: Various industries like banks, use the amount
Branch visits: difference between e-banking and conventional RI ELJ GDWD WR JHW WKH PD[LPXP NQRZOHGJH RI LQGLYLGXDO¶V
banking. Web and social media attachments: for effective needs, urban planning which helps in creating fruitful results
marketing. for the consumers as well as banks and also it will help in
Key Areas for Using Big Data: optimizing the standards of the services which was offered by
To generate the personalization and comfortability. Latest them. Hence we explain the concept of voluntariness as
credit card risk examination. Faster loan and credit examining the full context of big data to have effective results
management for customers to make the process simple and in almost every areas inside the banking industries.
fast as much as possible. Versatility: Several authorities like banking industries,
government body functioning aims at environment modelling,
III. BIG DATA TRAITS prevention of environment, cost reducing methods. So this
The term big data has been in discussion from years, various term can easily be understood as the flexible nature of the big
industries are now realizing that if they put a step forward in data various industries can refer it in various contexts.
gathering the information which they are having than by applying
IV. TOOLS AND TECHNOLOGIES
analytics they can get deeper insights as well as some valuable
information from it [4]. Earlier also before they came to understand
big data they are using little bit of basic analytics which are in the A. Archietecture, Frameworks and Tools
form of spreadsheets. The various benefits are found by the industries The basic framework in a big data analytics is very much
as use of big data analytics adds speed and efficiency in gathering the comparable to the earlier ways of analyzing the information.
deeper knowledge as well as full image of the consumer which The major difference which we found is in the way of
ultimately helps in increasing the growth rate of the industry. executing the raw information. As the big data is very vast
according to the definitions the examination takes place by
breaking the information into various segments and then

2019 International Conference on contemporary Computing and Informatics (IC3I) 223

processed between multiple nodes. In addition with it the e[WUDFWLRQRUZHFDQVD\WKDW³VXUYLYDOWRWKHILWWHVW´+HUHWKH
presence of open source platforms like hadoop or map reduce solution is referred as chromosomes which are able to
in the cloud generates the big data analytics in different combine and mutate. These chromosomes are obtained for the
domains [2]. The tools which are used for analyzing big data survival in the environment which was provided to them for
are complex in nature therefore we need to have a source the execution of each individual in the collection. Sometime
which have a variety of skills. The unitary part is the data itself. we also refer this method as evolutionary algorithm. To solve
All the information which we have obtained from various non-linear problem such algorithm are much more convenient.
sources weather structured or unstructured are entered together Machine learning: the analysis of big data could be easily
to solve the purpose of examining. At this point to the
done with the help of machine learning. This machine learning
information derived is in the raw state and hence we need to
natural environment helps the system to trace the
convert in the suitable reason as well. Data warehouse is
secondary approach in which all the information which was undiscovered insights from the vast data with the help of
stored and then aggregated and after that it will further move algorithms as here there is no requirement of the explicit
on the process of examination. But the data is unavailable for programs.
real time. The steps are taken to make the vast data cleaned and Principal component analysis: It is the widely known
was made fully ready. mathematical tool used for highly typical data analytics. It
reduces the size of the vast variety of the data after the data
was analyzed. This is the method which has been used in the
fields of face recognition, reduction of the information. The
conversion of this multilevel data into the reduced data sets
helps in getting more deep insights of the information
collected.

V. PROS AND CONS OF BIG DATA ANALYTICS IN

BANKING

A. Impact of Big Data Analytics in Banking

Fig2: Applied architecture in big data analytics[12] The biggest challenge which was faced by the industry is that
the customers does not want them to treat them as a person
After this stage in this general framework, various decisions
from the flock but they want them to consider themselves as
are taken with regard to data inputs measures, distributed
the separate party. The major issues are confirmation from
design, selection of tools and several analyzing tools. At the
long term hospitality from top valued individuals, maintaining
end four major applications in big data analytics are obtained.
the presence of vast variety of the customers, cross selling of
These comprises of queries, detailed reports, online analytic
the goods about what product is referred to whom, fraud
processing (OLAP), and mining of the data.
detection, application monitoring, etc these are the crucial
areas where it has to be at a great concern. Time to time
analytics comes in the surroundings here. This helps banking
B. Techniques For Analyzing Big Data in Banking
industry to obtain useful information about the consumers,
To analyze the big data there are various techniques available analyze fraud activities, helps in monitoring of applications,
to us which helps us to memorize, analyze, to make changes maintaining relationships between explanatory and predicted
and collect the big data to made this type of information diversities from earlier happenings and helps us to analyze
volume which is easy to control [9]. about the future.
Some of the are discussed below:
Cluster analysis: It is one of the way which is very common
statistical data analyzing technique which was used in several
areas which mainly includes machine learning, image analysis,
exploration of information. The best example to understand
cluster analysis is described as the gathering of customers
having similar groups for targeted business execution. These
kind of analytics is not supported as training information are
not used. It is generally used for data mining.
Data Mining: it generally uses the set of methods to identify
the patterns from inside the vast datasets by collecting Fig3: Future of banking [13]
methods from statistics as well as machine learning with the
help of data base management concepts. These also includes The ways in which big data analytics are useful are:
methods like association, classification, cluster analytics. 1. Fraud Detection: It is the major area of concern in almost
Genetic Algorithms: it is a method which is used in every industry and banking as well as financial firms, due to
optimization which is collected during the process of natural which it makes very expensive cost to them. For every single

224 2019 International Conference on contemporary Computing and Informatics (IC3I)

party it is very harmful because it can lead to extent the place 9. Customer Lifetime Value (LTV): It is basically defined
where one can loose their identity at the first segment as well. DV WKH SDUW\¶V OLIHWLPH YDOXH LH about how long will be the
The advancements in the technologies has cleared the way of period that banks can retain its party. Understanding about
the cyber criminals to make such frauds. who are the best consumer, making them best at various
2. Application screening: The advance prediction can prove ways[5] and if banks may get victory on them by preventing
useful to banks in examining the large volumes of applications their loyalty are some of the areas for the banks to concern.
even without the removal of valuable variables with the 10. Feedback management: Obtaining feedback is really
absence of delays as well as errors, even without becoming very important as by analyzing the feedback banks could
tiered it can work with full efficiency and steadiness. Hence know about the services which are required by the party as
the outcome is very much accurate as well as authentic for the well as maintain the extent of the bonding between them for
use. long term reference.
3. Customer acquisition & retention: Analytics can help in
B. Reasons For Banks to Adopt BDA
the examination for optimized targeting, which makes it
convenient for the banking industry so as to quickly identify As per the report which was given by various researchers they
about high value party who are likely to respond. According to proposed that only 21 percent of the banking industry has
the information from vast reports we concluded that in banks employed the use of big data analytics whereas other view is
about 10% there would be an increase in the party that they would likely to invest in this scheme after one year.
opportunities every year. It can also help banks by giving
insights that how the industry can retain their customers.
Customer retention is also the major area where banking
industry needs to be highly focused so as to prevent their
parties to do migration. Loyal parties must get rewards time to
WLPHDQGLQGLYLGXDO¶VDWWULWLRQPXVWEHPLQLPL]HG>@
4. Knowing customer buying habits: Tracing for the right
SURGXFW DQG WDUJHWLQJ LQGLYLGXDO¶V XVDJH LV WKH ELJJHVW
difficulties which was faced by the banks which these
industries are trying to cure. By the help of big data analytics
banks can gather various customer segments and alter it with
the highly specified, can send talorized data to the party which
automatically lead to high response ratio.
5. Cross-selling: effective cross selling of the goods can only
EHKDSSHQE\WKHH[DPLQDWLRQRISDUW\¶VEHKDYLRULQGLIIHUHQW
places where high diversity products are available. Which
certain goods are to be sold to which individual which can
helps to track the successful outcome about the working of the Fig 4: Real time execution of data in banking industry [12]
cross sellers which enables the banks to increase the growth
rate in customer profitability as well as somewhere build the
strong bond with the party. The main reasons for the industries to employ BDA
6. Collections: Banking industry has a vast number of technologies are:
customers who always deposit the payment in time and 1. To attain productivity inside their industries.
defaulters as well. This is very strategic situation that on 7RDFKLHYHFRQVXPHU¶VVDWLVIDFWLon as well as their growing
which customer we need to have the major focus. needs and the attitude of their spending nature.
7. Better cash/liquidity planning: This analytics could help 3. At last, to increase the growth cycle in the areas like total
banking firms to trace the earlier spending patterns and daily revenue as well as generation of profits.
keeps the check on the transactions which are made by the As we see the low implementation of BDA advancements in
LQGLYLGXDOZHDWKHURQOLQHRUDWUHVSHFWLYHEUDQFKHV>@$70¶V India at present ± a new trend was developed which was
provides a clear cut view of the party potential level. This recognized among the vast markets at Asia Pacific, we had
management could also help to tell about how much their concluded that from this enormous world various industries
LQFUHDVH LQ WKH FXVWRPHU¶V LQFRPH OHYHO DQG ZLWK UHVSHFW WR are making plans to have investments in the field of BDA in
this it can also tell which schemes are there in which they can coming future. As India is the place where economy is
invest. growing at very rapid rate, is in the operational state to
8. Marketing optimization: This type of analysis could help leapfrog different countries to help them to employ this BDA
banks or firms to plan marketing campaigns as well as method which can lead them to turn the raw facts or simply
programs and to make check on them steadily. By providing information into insights.
information of individual attitude as well as behavior, also it
helps to achieve 360 degree trait of the party.

2019 International Conference on contemporary Computing and Informatics (IC3I) 225

 4. Careful implementation of the tools like hadoop could be bit
difficult for the employees but if it comes in their hands than
the task could be simplified.

VI. CONCLUSION
The Big data analytics is continuously employing by various banking
industries in addition with several organizations across the world.
Energy must be consumed by the Indian banking industry so as to
make the services efficient which was provided by them and their
way of functioning as well. If the banking industry really want the
benefits to be obtained from BDA then they must have high necessity
to renew their old ways in approaching technology and begins to
derive new technologies in addition with the methods to meet the
Fig 5: Percentage of processing in banking [10] requirements as well. This project examined crucial areas of the
working and getting benefits from the BDA topologies. May be there
is the possibility at the time of using it which will enables us to know
C. Open Issues and Challenges for Indian Banks to Adopt Big the prediction when the Indian industries in banking would employ
Data Analytics BDA methodology. It also identifies the architectural frame works
The major issue which the banking industry is facing is what which are really very beneficial by examining the frauds which can
methods they can adopt to enter and analyze the high amount be performed during the period of online transactions and hence the
of the information gathered together which have new and vast experts can curb it as much quick as possible. To employ such
insights and in addition has variety as well. Banks are still in detections while performing online payments BDA takes help from
the position of exploring and practicing state about how BDA big data as well as machine learning so we can obtain effective
can be employed during the regular services offered by them. performance. At the time of transactions information are collected or
Although banks knew that the investment in this modern gathered, cleaned and important raw information is stored from each
party. It also proves helpful in the bankruptcy and also enables the
technology will definitely return its worth but still they are in individuals to give the view that the industry is going to be collapsed
deep confusion about the ways to calculate the advantages or not. Banking firms uses this big data which they called as the
from this source and future challenges as well. occupation of insights. Not only this, it can be extremely useful
among various fields like collecting raw data and by providing the
current status of market. Employing maximum correlations
observations which can vary according to the needs as well the
difficulties which are lying in front of the path which we follow. So,
we can conclude from here that if the banks are able to optimize the
resources to the fullest, could predict the hidden potentials and
various knowledge in vast areas so that banks can work on the way of
offering and providing the services to all the consumers.

REFERENCES

[1] N. Balasupramanian, Teresa. K. George, E. BeQ*HRUJH³$Q,QQRYDWLYH

Fraud Detection Framework for Secure E-Transactions using Big Data
$QDO\WLFV´(XURSHDQ-RXUQDORI6FLHQWLILF5HVHDUFK,661-216X /
Fig 6: Estimation of big data in banking industry[14] 1450-202X Vol. 122 No 3 May, 2014, pp. 342-349.
Some challenges are discussed below: [2] Dong Hyun Jeong, Caroline Ziemkiewicz, William Ribarsky and Remco
1. Banking industry in India would make sudden or steady &KDQJ³8QGHUVWDQGLQJ3ULQFLSDO&RPSRQHQW$QDO\VLV8VLQJD9LVXDO
move towards data culture which would not be so, easy as $QDO\WLFV7RRO´&KDUORWWH9LVXDOL]DWLRQ&HQWHU81&&KDUORWWH
individual see. fdhjeong, caziemki, ribarsky, 2009.
2. Banks need to trace the present sets of skills of each [3] Santaji Ghorpade, JayVKUHH*KRUSDGHDQG6KDPOD0DQWULULQJ³3DWWHUQ
employee and thus find out why there is a gap to employ BDA 5HFRJQLWLRQ8VLQJ1HXUDO1HWZRUNV´,-&6,7,661- 0975-9646, vol 2,
technologies as well as cater in same issue. No. 6, pp92-98, 2010.
3. The expenditure for the handling of BDA may be expensive [4] https://www.sas.com/en_us/insights/analytics/big-data-analytics.html
despite of this it would be a one-time investment and can [5] https://ieeexplore.ieee.org/document/07889917/
generate profits in the future [6] H. Mühlbacher, L. Dahringer, and H. Leihs, International Marketing: A
Global Perspective: International Thomson Business Press, 1999.
[7] https://ieeexplore.ieee.org/document/08342647/

226 2019 International Conference on contemporary Computing and Informatics (IC3I)

[8] https://www.newgenapps.com/blog/10-ways-predictive-analytics-help- [11] http://www.allerin.com/wp-blog/wp-content/uploads/2017/11/Top-5-
the-banking-sector sources-of-big-data.jpg
[9] https://sapinsider.wispubs.com/Assets/Articles/2013/October/SPI_Colu [12] http://www.ittoday.info/ITPerformanceImprovement/Images/2014-07
mn_Analytics RaghupathiFigure1.png
[10] https://thefinancialbrand.com/38801/big-data-profitability-strategy- [13] http://www.ronaldvanloon.com/the-future-is-bright-for-banking/
analytics-banking/ [14] https://thefinancialbrand.com/38801/big-data-profitability-strategy-
analytics-banking/

2019 International Conference on contemporary Computing and Informatics (IC3I) 227