Big Data is a concept that has become popular since 2012 to
express the exponential growth of the data to be processed.
These big data go beyond intuition and human analytical abilities. They require new tools to store, query, process and view information.
(Big) Data infographic - EnjoyDigitAll by BNP Paribas
1. by
(big)DATA?
Big Data is a concept that has become popular since 2012 to
express the exponential growth of the data to be processed.
These big data go beyond intuition and human analytical
abilities. They require new tools to store, query, process and
view information.
The masses of data
to be processed are
constantly increasing.
More and more, data
collection, analysis and
use must be done in
real time.
It's about
focusing on valid
and actionable
data.
The reliability of data is
threatened by declarative
behaviors (forms), by the
multiplication of data
formats and by bots
activity and false profiles.
VOLUME VELOCITY VARIETY VERACITY VALUE
We often understand the dimensions of Big Data thanks to the 5V :
The data are varied and not
always structured (Social
networks data for example).
90%
of data is
"unstructured"
DATA FLOWS DEVELOPMENT
Unlocking the Mysteries of Data
2013 : 28 875Go/sec
2018 : 50000Go/sec
1992 : 100Go/day
2002 : 100Go/sec
1997 : 100Go/hour
WWW
Structured DATA
Structured data are organized and classified information to facilitate their reading and processing.
Your customer databases are structured data.
Date of birth
Adress
Loyalty
Transactions
NameAmounts
Semi-structured data is an intermediate form. They're not organized in a complex way
that allows sophisticated access and analysis; However, some information may be associated with them,
such as metadata tags, which allow the addressing of the elements contained therein.
Unstructured data are not organized in a format that allows to be access and process
easily. In fact, few data are completely unstructured. Even elements often considered
unstructured, such as documents and images, are structured to some extent.
A Word document is generally considered as a set of unstructured data.
Semi / Non-Structured DATA
Products Reviews
Tweets
Likes
Images, etc...
90% OF THE CURRENTLY
AVAILABLE DATA WERE
CREATED THE PAST
TWO YEARS!
2000 20201970
Data analysis is the part of Data
Science that dissects raw data
by applying algorithms.
Data analysts proceed by
inference: from well-
known premises to new
conclusions in order to
improve systems and
decision-making.
ANALYTICS
DATA
WHAT TO DO WITH ALL THESE DATA?
We can distinguish, without losing the complexity of uses, two great potentials:
Linked to the exploitation of the
collected information in order to
understand a complex target or to
create corpus of information
improving the algorithms of AI...
INFORMATIONINFORMATION
Artificial
Intelligence
INSIGHT &
E-REPUTATION SEGMENTATION,
PROFILING,
TARGETING
BIG DATA
Linked to the data exploitation to
improve of the performances for
example with a global control
panel or by carrying out specific
actions of optimization!
PERFORMANCEPERFORMANCE
OPTIMIZATION
DASHBOARD
(tracking ROI, RTB programmatic...)
DATA
S
C
I
E
N
C
E
D
A
T
A
Designed by
by
Sources : Definitions-Marketing.com, Le Big Data au Quotidien - Vouchercloud.fr,
« United Nations Population Division » - Organisation des Nation Unies, Lexique - Nordnet.com
SimpliLearn.com, Wikipedia.fr, FlatIcon.com, « Données Semi-Struturées » - LeMagIT.fr