Page 42 - BANKING FINANCE APRIL 2017

ARTICLE

TOTAL FOR COMMERCIAL BANKS 'April 2011 'October 2015 INCREASE % INCREASE
Debit Card - No of Transactions (ATM+POS) 422.01 Mi 803.24 Mi 381.23 mi 90%
NEFT - No of Transactions (Incoming + Outgoing) 6.60 Mi 16.66 Mi 10.06 Mi 152%
Credit Card - No of Transactions (ATM+POS) 23.40 Mi 69.44 Mi 46.04 Mi 197%
NEFT - No of Transactions (Incoming + Outgoing) 29.72 Mi 229.20 Mi 199.48 Mi 671%
MOBILE - No of Transactions 1.08 Mi 32.08 Mi 31.40 Mi 2907%

Although banks were scaling up constantly to address this
issue of data explosion by adopting RDBMS, Data ware-
houses, Data Malls and CRM techniques, the readiness was
for representing the structured data related to customers,
products and assets. Because these data is unique and con-
sistent and coming from the main frame, quality about the
data has improved in the long run. But the issue of unstruc-
tured data that is creeping into the system through various Social Media Photos Audio & Videos
medium was left unanswered.
The large repositories of internal and external data which
What is Big Data? forms the Big Data in banks will help to uncover trends,
statistics, and other information for the decision making for
Let us take the example of unstructed data that is gener-
the fact more the information, better the decision will be.
ated in the banking arena in today’s environment. One such
But problem with Big Data is that because of the presence
data is the "log data" that is generated through various pro-
of both structured and unstructured data, it is difficult to
cesses in the system. We also have the sensor data which is
bring into an uniform format and do the analysis of data.
captured at the point of contacts.
According to Mr Robert LeBlanc, Senior VP, IBM going by
Still images/Video and Audio captured at the time of doing
the Big Data composition, following are the characteristics
an ATM transaction also forms data resource. Added to
of Big data, often called "V cubed".
these is the Social Media (FB, Twitter, etc) streams. Thus
any structured and unstructured data created can be
1. Variety - Which comes form complexity of multiple re-
brought into the ambit of "BIG DATA" and banks are gener-
lational and non-relational data types and schems
ating a lot of this Big Data on a daily basis.
2. Velocity - Derived from streaming data and large vol-
Big Data Source ume data movement and
3. Volume - From generation of data from different sources
and the resulting growth of Scale from terabytes to
zettabytes

Because of the above characteristics of Big Data, traditional
approach of data analysis will not work as traditional ap-
proach can be applied only on Structured data where re-
peatability is there and source is stable. The other side of
traditional approach, as we all are aware, is it cannot sup-
port variability and is difficult to scale. Having said that, it
Transaction Data Log Data E-mail doesn't mean that Big Data Analytics will replace a rela-
(Hierarchical) tional database management system (DBMS) or a tradi-

42 | 2017 | APRIL | BANKING FINANCE

Insurance Training Centre © Call 09883398055 / 09883380339