Professional Documents
Culture Documents
02
Data Science |
Transforming the ICT domain
Contents
Foreword 4
Message from Aegis School of Business 5
Data Science in ICT 6
Benefits Realized through Data Science 11
Pointers for ICT companies 15
The Future 17
Concluding Remarks 20
About Aegis 21
Acknowledgments 22
Contacts 22
References 22
03
Data Science | Transforming the ICT domain
Foreword
Information and Communication applications. The rapid development in
Technology (ICT) as a sector has seen a the domain of Internet of Things and
phenomenal growth in the past decade. mobile devices has helped us realise once
These companies have access to a great highly ambitious projects of Autonomous
amount of data generated by digital Cars and Smart Homes. The demand for
footprints left behind by customers which people to work in this interdisciplinary
is increasing exponentially. ICT as a sector domain of Data Science and ICT is an
has gone an overhaul due to this boom upward trend, far outgrowing the supply,
of data generation. Data storage capacity which defines the need for skilled data
of devices across the globe to store and scientists with ample knowledge about
receive data over network and compute emerging information technologies
has increased more than ten-folds in a and the ability to implement business
very short duration. All this has driven solutions effectively.
adoption of new ICT technologies like
Cloud Computing, Internet of Things, Having set the complexities of data
etc. This certainly demonstrates the science, application of these technologies
emergence of Data Science to convert has become simple and one does not
data into business insights. require high end development skill in
order to adopt data-science skills. With
The credit for optimization achieved the help of easily available libraries
by these technologies in managing and user-friendly tools, business users
data using distributed storage and can develop and adopt data science
computation capabilities goes to technologies. Easy access to web-hosting,
advancement in Data Science and its cloud computing and AI tools has enabled
close relationship to developments in the business users to focus more on business
ICT Sector. The collaboration between logic and achieve quicker business
industry and academics has been a insights and outcome.
driving force behind the exponential rise
in the usage of ICT devices. Both these
technologies, Data Science and ICT, have
formed an interdependence which helps
both these domains to aid development.
04
Data Science |
Transforming the ICT domain
05
Data Science | Transforming the ICT domain
06
Data Science |
Transforming the ICT domain
Technology Pillars
Statistics: Many critical academics and
realise data science goals and support
for the same has made Python and R,
“The exponential growth
scholars believe that data science is just a both – popular languages in the market. in the adoption of digital
fancy term for statistics, and that statistics
is at the very core of data science study.
Emergence of open source licensing
platforms like GitHub has helped to grow
technologies by new
All these critiques are true to say that the number of contributors which has age and traditional
statistics forms a very important pillar in
the field of data science, providing the
led to development of various packages
and API Frameworks for ease of use by
organizations presents
very basic as well as complex metrics to the academia as well the industry to numerous opportunities
solve and evaluate the analysed set of
data. Some commonly and everyday used
implement Data Science Solutions in their
businesses.
for people looking to build
metrics include type of data (Continuous a career in this space.
vs Discrete), statistical distributions
(Poisson, Binomial, etc.), Probability and
Domain Knowledge: Domain knowledge
is used all the time in Data science
Notwithstanding concerns
Cumulative density functions and accuracy applications (sometimes without knowing that automation, artificial
analysis using ROC curves. Some statistical
theorems form the base for various data
that you are doing it). A good example
is feature extraction, how do you know
intelligence and machine
science algorithms like Bayes Theorem, that these features are important for the learning will result in job
KNN Algorithm and Bagging/Bootstrap
aggregating. Thus, we can say that
model which you are building? For e.g.
Internet Speed may be good features in
loss, there are whole new
knowledge of statistics is very important knowing about how much time a customer areas of work where we
from the point of view for real world
application of data science.
spends on online content streaming.
Domain Knowledge helps to identify which
are witnessing an acute
data set make sense to business and how shortage of skilled people.
High Level Programming Language: they are going to be consumed. All that is
Python and R, are two prominent open
source, high-level versatile programming
needed to build a data science model is
a dataset which consists of examples or
The edge of innovation
languages used for almost all purposes records in rows and attributes or features is an ever expanding
over the breadth of the industry. The
open source community is very robust
in columns. A Data Science model needs
a response feature which is what would
line. Acquiring analytics
which has given rise to a variety of inbuilt be predicted when the other attributes and data science skills
packages which makes these languages
very easy-to-use and an important
are known only with the help of domain
knowledge.
across the spectrum of
tool for data science. The growing business, mathematics
abundance of libraries and packages to
and technology will enable
the workforce of today to
stay relevant, capitalize on
the opportunities and ride
the growth wave.”
Rajan Sethuraman,
Chief People Officer at Latent View
Analytics
07
Data Science | Transforming the ICT domain
01
Cloud Computing
02
Internet of Things (IoT)
08
Data Science |
Transforming the ICT domain
03
Artificial Intelligence
04
Networked systems
05
Mobile services
09
Data Science | Transforming the ICT domain
63.9
61.8
59.7
57.5
55.4
53.2
50.5
48.8
46.9
44.5
41.6
38.2
Avik Sarkar - Head, Data Analytics Cell, NITI Aayog at Government of India
10
Data Science |
Transforming the ICT domain
11
Data Science | Transforming the ICT domain
developmental insights on the importance to companies as they seek to Cyber Security: The biggest advantage
population, since the service with the understand the demands placed on their with data science is that it can, indeed,
widest coverage and greatest uptake networks by the use of popular Over the assist security analysts in detecting
and popularity is the cellular service. Not Top (OTT) services. actual threats more quickly and allow
surprisingly, the data for development organizations to act proactively. This
initiatives have mainly drawn on mobile- Product Recommendation: is achieved through in-depth historical
network data rather than on those from Data Science-developed product analyses of security data.
fixed-line telephone operators or ISPs. recommendation systems learn
behavioural shopping patterns, such Machine data is not just logs, but
Competitive Advantage for Early as purchasing similarities between comprehensive records of behavior of
Starters for adopting Data Science customers or relatedness of search end-users, server, networks, applications,
Customer Profiling: Thanks to the items, to predict customers’ preferences transactions and mobile devices.
phenomenal progress on the technology towards new items. These prediction- It’s not limited to API data, machine
front, ICT companies can capture a wide based recommendations lead to higher configurations, message queues, events,
gamut of behavioural data about their sales revenue by exposing customers CDR (call detail records), IoT (Internet of
customers. These profiles include details to additional products of interest and Things) data, sensor data from industrial
about customers’ mobility patterns, encouraging upgrades to more expensive machines, automation and many others.
social network activity and personal products Consequently, in cyber security, machine
preferences. Collectively, these digital data is useful for fraud detection, artificial
breadcrumbs enables the companies Pricing Strategies: Data Science models intelligence and recommendations. Data
to segment their customers based on are commonly used to develop optimal Science also detect changes over time
a variety of parameters. Depending on pricing strategies to maximise profits. that render network behavioral profiles
the geographical region, there may be Dynamic price optimisation is a revenue of normal vs. abnormal traffic without
different privacy and data regulations management tool widely used in retail, manual intervention.
governing the manner in which the automotive, mobile communication and
telecom companies can gather or use electricity industries, and is generated Gaining Competitive Advantage: In
this data. This affects the working of using data on variability in customer studies conducted up to five years ago,
the operators to gather insights by preferences and buying patterns. researchers found that companies
behavioural profiling to a great extent. that use data science in their decision-
New Business Lines: It is natural for making were 5% more productive and
Network Planning and Management: operators to leverage the data they 6% more profitable than competitors.
ICT companies can optimize the network hold for better insights to increase their 17 other studies showed that firms with
routes and improve their Quality of revenue streams. The customer insights these capabilities were also five times
Service by continuously analysing the obtained paves the way for creation as likely to make decisions faster than
network traffic in real time. The use of of new business lines, either through competitors and three times as likely to
real-time Deep Packet Inspection (DPI) innovation or by partnering with other have faster execution on those decisions.
enables gather details of the current businesses, including credit-scoring and
traffic volumes, including the geospatial other financial services. One example
distribution of demand, and to manage is of a US-based big data start-up which
their network connections effectively obtains data from telecom operators
using optimal resource allocation.2 and financial firms to build customer
portfolios and in turn evaluate the
Operators can adapt their resource creditworthiness.2 Cross-promotions
allocation to ensure that more resource with brick-and-mortar businesses are
is allocated in high-revenue regions a potentially high-growth area in which
where most active customers reside by the detailed mobility profiles available to
utilizing the geospatial information from operators are leveraged.
their devices. This is a niche area of great
12
Data Science |
Transforming the ICT domain
“In the age of digitalization Revenue and Growth Benefits buying patterns. Data science models
Businesses can use data insights derived are increasingly necessary to handle the
and artificial intelligence using data science to improve customer volume of telecom data in real-time. By
how can we avoid engagement by better identifying, using Data science to make use of all
understanding and responding to available data and generate customized
dehumanization of society customers. For example, speech and text offers for the right customer at the right
by machines? analytics techniques can be applied to time, telecom companies can increase the
document live customer interactions to probability of sales and generate higher
We come from the stance follow up on potential sales leads. Data revenue.3
that algorithms with no science can improve efforts at customer
segmentation to predict the most Capital Savings
base in human thinking profitable, or most risky customers. In There are many ways data science help
or solutions that derogate the insurance industry, predictive models businesses make better use of their
that identify high risk customers help physical assets and budgets. Optimizing
and enslave people are insurance businesses minimize losses different stages of the production
not intelligent in human and develop appropriate premiums. For process, from inventory management
online businesses with large product to quality control can deliver substantial
sense. We believe that ranges, machine-learned ranking (MLR) savings to ICT companies. These
only cognitive algorithms can improve product search using characteristics are inherent for a leader in
personalization based on a user’s search data science.
that traceably emulate real and previous purchasing history, which
good human heart logic means customers can find and purchase Other ICT companies can certainly
the right products faster. use predictive methods of analysis
and moral norms should for optimization to save on capital
interact with us humans. The powerful combination of business expenditure by following leaders. Models
expertise and data science drives better can be trained to utilize user usage data
Because our mission strategic revenue growth decisions. Data and accordingly optimize the placements
as the scientific society, Analysis models are commonly used of towers and laying of network lines
to develop optimal pricing strategies saving the companies a lot in capitals
individuals and companies to maximize profits. Dynamic price savings. Also, study of the usage can
is to enable progress optimization is a revenue management help the companies to take decision
tool widely used in retail, automotive, with regards to either setting up own
of values that make us mobile communication and electricity services or rent the components from
human.” industries, and is generated using data on
variability in customer preferences and
Nikola Sucevic
Algorithms/Analytics - Data Science
Head at Reliance Jio
13
Data Science | Transforming the ICT domain
outside. We see that many companies higher value activities. Data Science
provide services to customers from their techniques aid in the development of
competitors so as to save on capital applications that automate tasks and
costs and get the same benefit for their augment existing processes, leading to
customers in return. Thus, data science productivity improvements and cost
can play a vital role in optimization of savings for ICT companies. Many routine
network planning and service delivery for internal processes can be done quicker
these telecom companies.3 and more systematically by machines
than people. Models can provide
Time and Efficiency Benefits preliminary structure to raw data,
Businesses can achieve significant time saving people from performing routine
and efficiency benefits by using Data tasks that are highly time consuming,
science applications to cut down on letting these high-skilled workforce to
costs and shift human resources to concentrate on high productivity tasks.3
Figure 2: Early starters extensively uses different digital platform for competitive Advantage
Advertising Transaction
Search Engine
Google
Application Stores
Apple, Play Stores
Social Media
Ride sharing
Classified Digital
Platforms
Market Place
Fermium
Pay as you go
Subscriptions Infrastructure Amazon
Digital Media Netflix
Repository GitHub
14
Data Science |
Transforming the ICT domain
15
Data Science | Transforming the ICT domain
Vijay Gabale,
Co-founder and CTO at Huew
16
Data Science |
Transforming the ICT domain
The Future
How is data science going to shape up content delivery network (CDN), which
the future business strategies? Let us is tailored to one specific application:
understand how Netflix and Amazon are delivering internet TV to its members
using content streams service to drive around the world. This system alone
revenue. is responsible to serve 100% of Netflix
content, over 125 million hours every day,
Understanding data science impact on to 100 million members across the globe!4
content delivery network
Have you ever wondered where your Such online content providers are the
video comes from when you watch future of content consummation, all
Amazon Prime, Netflix or Jio TV? Netflix thanks to the ease of use and on-the-go
serve video streams out of our own access to all your content from multiple
17
Data Science | Transforming the ICT domain
devices. Internet and advances in the companies, which use the humongous Netflix has more than a 100 million
communication sector has certainly amount of data they have in the best customers worldwide and it actively uses
helped such companies to flourish possible way to improve their services the data generated by these users to help
and garner a huge user-base for their and generate more revenue every day. them take crucial decisions.5 Keeping a
services. But, it is not just the ease of track of the content watched, completion
access that has led to the popularity of The core job of data science here also is rates of a particular TV show, age
content streaming service providers; one to gain insights into the customers, to demographics and region-wise indexes,
of the highly evolving field of data science optimize and deliver a better product. Netflix takes crucial decisions regarding
and its pro-active use is at the helm of the Data science enables these businesses buying licenses and production of Netflix
growing traffic of consumers that these to make informed decisions and improve original shows. For example, if Netflix
service provides witness. It won’t be their services considerably relying on the can derive from data that more than
wrong to say that Netflix and Amazon quantitative aspect. 60% of users watched a particular show
are two completely data-driven to completion, it might think to revive
or restart the TV show again with a new
season.
18
Data Science |
Transforming the ICT domain
and the dynamic electronics ecosystem. models that perform dynamic budget a previously time consuming task that
Each of these factors greatly affects the allocation. The model is trained on was prone to human error to be made
way traffic is served from a network historical data and provides suggestions more efficient and accurate. This allows
location, also what hardware systems will of how businesses should allocate their for analysis of higher quality satellite
be compatible and effective in the long marketing budget between different images and provides information on
run with respect to the dynamic nature of campaigns. cloud location so satellite systems can be
changes. reprogrammed as needed.
As these budget allocations are updated,
The major data science challenge is the model is retrained on incoming Airbus also uses data science to extract
to combine these various factors into data and can continue to optimize its information from satellite images for
medium- and long-term forecasts. The recommendations. This translates into big data-style analysis, for use in other
work involves a combination of demand real cost savings for businesses. Kofera’s applications, including for agricultural,
forecasting, system modelling — combining clients can save over 15% on marketing engineering or environmental purposes.
all complex factors together to build costs by using campaign monitoring Prior to using automated analysis
an efficient performance model — and systems which optimize campaigns every techniques, data scientists had to
resource analysis to identify domains of 15 minutes instead of weekly or monthly. develop rule-based algorithms to extract
over and under-utilization now and in the information in a geometric way; for
near future continuously. These benefits are quickly realizable too – example specifying that buildings are
benefits from cleaning data and using NLP likely to be rectangular. By picking up
Use Cases6 are virtually instantaneous, and benefits on patterns on their own, data science
Kofera: Digital Marketing in Asia from marketing optimization are realized algorithms are an improvement on this,
Kofera uses data science learning to help in as little as one to two months. and more accurate.
e-commerce and online retailers build,
optimize and monitor their advertising and Airbus: Operational Efficiency Airbus expects future applications to
marketing campaigns. Based in Indonesia, Airbus Defense and Space utilizes data continue to achieve improvements. They
Kofera’s clients range from small and science in a number of applications. One have already seen a significant reduction
medium businesses (SMBs) to large of these is the detection and correction of in the time to build algorithms, from
businesses with over a million products. satellite images with imperfections such several years down to a few weeks, and
as the presence of clouds. For example, it this reduces time to trial so that new
Data Science driven data-analytics is the can be challenging to detect the difference applications can be developed.
future of marketing optimization. One between clouds and snow on images
example involves creating predictive by eye. Data Science techniques allow
19
Data Science | Transforming the ICT domain
Concluding Remarks
The potential business outcomes •• Process Digitization: Ensure existing
delivered through data science can be processes and data collection channels
an attractive prospect for ICT sector. are digitized and creating data inputs
However, there are considerations that that are relevant for applying data
organizations need to recognize and science.
evaluate to maximize the returns from •• Data Lake: With the exponential
investment. As with any technology, the increase in various type of data
successful development and deployment (internal data, external data, partner’s
of Data Science applications within an data, competitor data, business
organization requires various capabilities process, social data, and people data)
and skills. In developing technology and organization need to build data lake to
capability, there are considerations in solve various data science problems.
terms of business outcome, Data Lake, The Data Lake excels at utilizing
technology architecting and process the availability of large quantities
digitization. of coherent data along with deep
•• Business Outcome: Translating a learning algorithms to recognize items
business problem into a testable of interest that will power real-time
hypothesis – Defining how to decision analytics.
measure successful outcomes from
tests – Establishing baselines to Implementation and change
enable assessment of incremental management is an important final step
benefits delivered by data science in a Data Science project. Data science
use cases – Quantifying success, and transformations are not short term
translating insights into format that is endeavors, and subsequently require
understood by less technical colleagues organization-wide transformation.
– Commercially minded (cost vs Effective implementation is crucial to
benefit etc.). realizing the full benefits of Data Science
•• Technology Architecting and applications; and constant support of
Implementation: Technology engine the organization and executives are an
which collect, consolidate and nurture important part of such endeavors.
data analytics on real time/scheduled
jobs for applying data science.
20
Data Science |
Transforming the ICT domain
About Aegis
Aegis School of Business, Data Science, Aegis has also initiated worlds largest
Cyber Security and Telecom was founded innovation award for Telecom, Internet,
in the year 2002 with support from Media and Edutainment (T.I.M.E) and
Bharti Airtel to develop cross functional Social, Mobility, Analytics and Cloud
technology leaders. In 2015 Aegis and (SMAC) know as Aegis Graham Bell
IBM collaborated to launch, India's first Awards for developing an ecosystem for
Post Graduate Program (PGP) in Data fostering innovation in India. This award
Science, Business Analytics and Big Data in organized with support of Cellular
and later in 2017 PGP in Cyber Security. Operators Association of India (COAI);
These programs are jointly certified and Convergence India; Deloitte and Telecom
delivered by Aegis School of Business Centre of Excellence (TCOE).
in association with IBM. IBM has set up
high end Business Analytics and Cloud For more information, please visit
Computing Lab at Campus. Also Aegis www.BellAward.com
and NVIDIA partnered for Deep Learning
and applied AI courses. Aegis is no. 1 In 2017, Aegis launched Data Science
School of Data Science and among top Congress to create a vendor-neutral
5 in Business Analytics. Aegis takes up platform where different stakeholders
various industry projects, research and including policy makers, industry,
consulting assignments in the field of experts, data scientist, CIO, decision
data science under its initiative "Data makers can share their knowledge, best
Science Delivered" and "Data Science for practices, innovations, products, uses
social good", and helping organizations cases and establishing a dialogue among
for devolving skills on data science, ML, the practitioners, users and tech vendors.
DL, Big Data, Analytics etc.
For more information, please visit
For more information, please visit: www.DataScienceCongress.com
www.Aegis.edu.in &
www.mUniversity.mobi/Aegis
21
Data Science | Transforming the ICT domain
Acknowledgments
Hemant Joshi Payal Agarwal
Senthilvel Kaliyamurthy
Contacts
Aegis School of Business, Data Science, Cyber Deloitte
Security & Telecommunication 7th Floor, Building 10, Tower B,
7th Floor, CETTM, Technology Street, Hiranandani DLF Cyber City Complex,
Gardens, DLF City Phase - II
Powai, Mumbai - 400076 Gurgaon, 122 002, India
Telephone: +91 22 2570 2815 Telephone: +91-0-124 679 2396
Website: www.aegis.edu.in e-mail: inideas-tmt@deloitte.com
e-mail: info@aegis.edu.in Website: www.deloitte.com/in
References
1
What Does Big Data Analytics Need from ICT to Develop? Feb 2016, See: http://blog.agroknow.com/?p=4807
2
Measuring the Information Society Report 2014, UN Global Pulse, See: http://www.unglobalpulse.org/sites/default/files/Pages%20
from%20MIS2014%20-%20Big%20Data%20Chapter.pdf
3
Deloitte Access Economics – Business Impact of Machine Learning
4
How Data Science Helps Power Worldwide Delivery of Netflix Content, May 2017, See: https://medium.com/netflix-techblog/how-
data-science-helps-power-worldwide-delivery-of-netflix-content-bac55800f9a7
5
Number of Netflix streaming subscribers worldwide from 3rd quarter 2011 to 4th quarter 2017 (in millions), Statista, See: https://
www.statista.com/statistics/250934/quarterly-number-of-netflix-streaming-subscribers-worldwide/
6
Deloitte Access Economics – Business Impact of Machine Learning
22
Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK
private company limited by guarantee (“DTTL”), its network of member firms,
and their related entities. DTTL and each of its member firms are legally
separate and independent entities. DTTL (also referred to as “Deloitte Global”)
does not provide services to clients. Please see www.deloitte.com/about for a
more detailed description of DTTL and its member firms.
No entity in the Deloitte Network shall be responsible for any loss whatsoever
sustained by any person or entity by reason of access to, use of or reliance on,
this material. By using this material or any information contained in it, the user
accepts this entire notice and terms of use.