�ݺ�ߣ

Findwise seminar 29.2.2012

Trends in analytics
Anders Quitzau - innovationschef

© 2012 IBM Corporation

What is happening out there?

2

We
are
ushering
in
a
new
wave
of
innova.on

Age 6th
Wave
Age
of
IT
&
of
Oil,
Cars Telecom
Age
and
Mass

of
Steel,

Produc2on 5th
Wave
Electricity
and
Heavy

Engineering 4th
Wave
Smarter
Innova.on

Age

of
Steam
3rd
Wave
The

and
Railways
Products
Industrial
§ Instrumented,

Revolu2on 2nd
Wave interconnected,

and
intelligent
1st
Wave
§ Building
blocks

for
a
smarter
planet

§ Sustainability

1770 1830 1875 1920 1970 2010

*Source:

“Next
Genera2on
Green:
Tomorrow’s
Innova2on
Green
Business
Leaders”,
Business
Week,
Feb
4,
2008

As the planet gets smarter, the information explosion and rapid
change create new challenges for CIOs

1.2 ZB 44x
Projected growth rate for digital
Size of the world's total digital
content in 20101 data from 2009 to 20201
(1 zettabyte = 1 billion terabytes)

22 billion 64 EB/mo
Number of devices connected The volume of global Internet
to the Internet by 20202 traffic expected by 20143
(1 exabyte = 1 million terabytes)

Source: 1The Digital Universe - Are You Ready?, John Gantz and David Reinsel, IDC, May 2010 2Internet Connected Devices About to Pass the 5 Billion
Milestone, IMS Research press release, August 16, 2010 3Cisco Visual Networking Index (VNI) : Global Mobile Data Traffic Forecast Update,
2009-2014, February 2010

Predictive analytics, big data, social analytics, virtual assistants on
Gartners Hypecycle

6 © 2012 IBM Corporation

Analytics is a major trend that will impact all business processes and
transforming the way decisions are made
Types of Analytics Source: Forrester

Trends in Analytics
From To
Offline / Back office Embedded / Realtime
Detailed Reports Dashboards
Historical Predictive
Structured Unstructured
Behind Firewall Cloud / Mobile


The key will be to leverage information and analytics to enable
informed, real-time decisions at the point of impact
The breakaway
approach

Sense and Predict and
respond act

Instinct and Real-time,
Lack of insight intuition fact-driven

Skilled analytics
experts Everyone
Inefficient access
Point of
Back office
impact
Inability to predict
Automated Optimized

Customer relations will require near real-time information and the
ability to change on the fly, which will require innovation.
– Energy and Utility CIO

We are crossing a threshold in our ability to produce information that
delivers insight and drives action

From To
Disease treatment Disease prevention

Intelligent highways
Traffic management
and cars

Financial crime Fraud detection
and fraud and prevention

Informed exploration
Speculative oil exploration
and targeted drilling


Findwise seminar 29.2.2012

Big data - analytics
Anders Quitzau - innovationschef


Every Smarter Planet Solution has Big Data and Needs Big Analytics

Smarter Planet

Faster Decisions Deeper Insights
Real-time Awareness Predictive Models
Reactive Analytics Deep Analytics
Data in Motion Data at Rest
Big Data

Smarter Computing

Integrating new approaches such as Big Data will unlock
new insights.

Traditional Approach New Approach
Structured, analytical, logical Creative, holistic thought, intuition

Structured Unstructured
Repeatable Exploratory
Linear Iterative
Monthly sales reports Brand sentiment
Profitability analysis Product strategy
Customer surveys Maximum asset utilization


How will Content Analytics be like in
2020?


New “Big Data” Brings New Opportunities, Requires New Analytics
Exa
Homeland Security
600,000 records/sec, 50B/day
Up to
10,000 1-2 ms/decision
Times 320TB for Deep Analytics
Peta
larger
Data Scale

Telco Promotions
Tera 100,000 records/sec, 6B/day
Data at Rest

10 ms/decision
270TB for Deep Analytics
Scale
Data

Giga

DeepQA
100s GB for Deep Analytics
Traditional Data 3 sec/decision
Mega Warehouse and
Business Intelligence
Data in Up to 10,000
Motion times faster
Smart Traffic
Kilo 250K GPS probes/sec
630K segments/sec
yr mo wk day hr min sec … ms µs 2 ms/decision, 4K vehicles
Occasional Frequent Real-time
Decision Frequency

Big Data Systems Require a Data-centric Architecture for
Performance

Old Compute-centric Model New Data-centric Model
Manycore FPGA

out
t put
in pu Massive Parallelism
Persistent Memory

Flash Phase Change
Data lives on disk and tape Data lives in persistent memory
Move data to CPU as needed Many CPU’s surround and use
Deep Storage Hierarchy Shallow/Flat Storage Hierarchy

Largest change in system architecture since the System 360
Huge impact on hardware, systems software, and application design

Watson - IBMs DeepQA
- den største IT-revolution siden transistoren


From the final match......


From battling humans at Jeopardy! to transforming business

To compete at Jeopardy!, Watson-like capabilities can
humans and computers need to: now be applied to your business:

Manage massive amounts of
Tap into a broad, open domain of
data from business operations
clues
and an instrumented world

Deal with structured and
Parse complex human language unstructured data to gain
meaningful insight into your
business

Be highly precise at finding the Have a high level of confidence
answer and put a confidence in analytics to make key
behind that answer business decisions

Do all of that in under three Do it at the speed of business, in
seconds real time

Watson’s analytics are more than search

§ Web search returns a ranked list of
‘possible’ web pages containing the
requested data
‣ Search engine results are based on
popularity and page ranking;
‣ User must still analyze results – sift
through a web page -- to find the best
answer.

§ Watson’s analytics understand the
structure and wording of the question
asked
‣ Finds a specific answer;
‣ Ranks its answer and provides a level of
‘confidence’ that this is correct based on
previous experience.

§ Watson answers ‘natural language’
questions
‣ May contain puns, slang, jargon and
acronyms that must all be evaluated.


Announced Aug 18, 2011 - IBM has developed prototype of
‘brainlike’ cognitive chip


Future Systems – The Learning Paradigm

Society Nature Institutions Archives

(Natural Interfaces)
Training and Learning Engines
To Build Models and Define Insight Policy Engine
Active Learning Business, Legal
and Ethical Rules

Verification Engines Hypothesis Engines
(e.g. Simulations) To Understand and Plan Actions

Outcome Engine
Actuation and Validation

60

Year 2020?


Prepare yourself and your organisation for 6.th wave!


�ݺ�ߣ

Trends in content analytics

More Related Content

Trends in content analytics