1. Big Data, IoT, Analytics & Foresight:
Natural Language Generation, Systems of Insight & Deep Learning
February 2018
Suresh Sood, PhD
@soody,
suresh.sood@uts.edu.au
linkedin.com/in/sureshsood
2. Areas for Conversation
Foresight
Data Science Innovation (s) – IoT & Big Data
Democratisation of big data
Gartner & Forrester Trends
Natural Language Generation
Systems of Insight
Deep Learning
4. 2017 Hype Cycle for Data Science and Machine Learning,
29 July, http://www.gartner.com/document/3772081
Gartner (2017)
Strategic Predictions for 2017 and Beyond, research note
14 October, http://www.gartner.com/document/3471568
By 2020-22 :
100 million consumers shop in augmented reality
30% of web browsing sessions without a screen
Algorithms positively alter behavior of over 1B
Blockchain-based business worth $10B
IoT will save consumers/businesses $1T a year
40% of employees cut healthcare costs via fitness tracker
Smart Data Discovery Will Enable New Class of Citizen Data Scientist
“With the addition of NLG [Natural Language Generation], smart data discovery platforms automatically present
a written or spoken context-based narrative of findings in the data that, alongside the visualization, inform the
user about what is most important for them to act on in the data.”
Gartner, 29 June, 2015
6. Vignettes in the two-step arrival of the internet
of things and its reshaping of marketing
management’s service-dominant logic
Woodside & Sood
Journal of Marketing
Management Volume 33, 2017 -
Issue 1-2: The Internet of Things
(IoT) and Marketing: The State of
Play, Future Trends and the
Implications for Marketing
7.
8. Statistics, Data Mining or Data Science ?
• Statistics
–precise deterministic causal analysis over precisely collected data
• Data Mining
–deterministic causal analysis over re-purposed data carefully sampled
• Data Science
–trending/correlation analysis over existing data using bulk of population i.e.
big data
–Extraction of actionable knowledge directly from data through a process of
discovery, hypothesis, and hypothesis testing.
Adapted from: NIST Big Data taxonomy draft report :
(see http://bigdatawg.nist.gov /show_InputDoc.php)
9. Useful References Big Data
• NIST Big Data interoperability Framework (NBDIF) V1.0 Final Version (September 2015)
Big Data Definitions: http://dx.doi.org/10.6028/NIST.SP.1500-1
Big Data Taxonomies: http://dx.doi.org/10.6028/NIST.SP.1500-2
Big Data Use Cases and Requirements: http://dx.doi.org/10.6028/NIST.SP.1500-3
Big Data Security and Privacy: http://dx.doi.org/10.6028/NIST.SP.1500-4
Big Data Architecture White Paper Survey: http://dx.doi.org/10.6028/NIST.SP.1500-5
Big Data Reference Architecture: http://dx.doi.org/10.6028/NIST.SP.1500-6
Big Data Standards Roadmap: http://dx.doi.org/10.6028/NIST.SP.1500-7
Note: Version 2 drafts available at https://bigdatawg.nist.gov/V2_output_docs.php
Use Case V2.0 https://bigdatawg.nist.gov/usecases.php - Request for Public Comments (Deadline: September 21, 2017)
• Apache Spark 2.1.0 Documentation
Machine Learning Library (MLlib) Guide http://spark.apache.org/docs/latest/ml-guide.html
GraphX Programming Guide http://spark.apache.org/docs/latest/graphx-programming-guide.html
SparkR (R on Spark) http://spark.apache.org/docs/latest/sparkr.html#sparkdataframe
Spark SQL, DataFrames and Datasets Guide http://spark.apache.org/docs/latest/sql-programming-guide.html
10. Data Science Innovation
Data science innovation is something
an organization has not done before or
even something nobody anywhere has
done before. A data science innovation
focuses on discovering and using new
or untraditional data sources to solve
new problems.
Adapted from:
Franks, B. (2012) Taming the Big Data Tidal
Wave, p. 255, John Wiley & Son
Data Science Algorithms
Companies are reimagining Business
Processes with Algorithms and there
is “evidence of significant, even
exponential, business gains in customer’s
customer engagement,
cost & revenue performance”
Wilson, H., Alter A. and Shukla, P. (2016),
Companies Are Reimagining Business Processes
with Algorithms, Harvard Business Review,
February
11. Variety of Data Types & Big Data Challenge
1.Astronomical
2.Documents
3.Earthquake
4.Email
5.Environmental sensors
6.Fingerprints
7.Health (personal) Images
8.Graph data (social network)
9.Location
10.Marine
11.Particle accelerator
12.Satellite
13.Scanned survey data
14.Sound
15.Text
16.Transactions
17.Video Big Data consists of extensive datasets primarily in the characteristics
of volume, variety, velocity, and/or variability that require a scalable
architecture for efficient storage, manipulation, and analysis.
. Computational portability is the movement of the computation to the location of the data.
12. • The data collected in a single day take nearly two million years to playback on an MP3 player
• Generates enough raw data to fill 15 million 64GB iPods every day
• The central computer has processing power of about one hundred million PCs
• Uses enough optical fiber linking up all the radio telescopes to wrap twice around the Earth
• The dishes when fully operational will produce 10 times the global internet traffic as of 2013
• The supercomputer will perform 1018 operations per second - equivalent to the number of stars in
three million Milky Way galaxies - in order to process all the data produced.
• Sensitivity to detect an airport radar on a planet 50 light years away.
• Thousands of antennas with a combined collecting area of 1,000,000 square meters - 1 sqkm)
• Previous mapping of Centaurus A galaxy took a team 12,000 hours of observations and several
years - SKA ETA 5 minutes !
To the scientists involved, however, the SKA is no testbed, it’s a transformative instrument which,
according to Luijten, will lead to “fundamental discoveries of how life and planets and matter all came
into existence. As a scientist, this is a once in a lifetime opportunity.”
Sources: http://bit.ly/amazin-facts & http://bit.ly/astro-ska
Galileo
Square Kilometer Array Construction
(SKA1 - 2018-23; SKA2 - 2023-30)
Centaurus A
15. 15
Sherman and Young (2016), When Financial Reporting Still Falls
Short, Harvard Business Review, July-August
Sood (2015), Truth, Lies and Brand Trust The Deceit
Algorithm,
http://datafication.com.au/
New Analytical Tools Can Help
16. 16
Deception Algorithm
(1) Self words e.g. “I” and “me” – decrease when someone
distances themselves from content
(2) Exclusive words e.g. “but” and “or” decrease with fabricated
content owing to complexity of maintaining deception
(3) Negative emotion words e.g. “hate” increase in word usage
owing to shame or guilty feeling
(4) Motion verbs e.g. “go” or “move” increase as exclusive words
go down to keep the story on track
17. Language on Twitter Tracks Rates of Coronary Heart
Disease, Psychological Science, January 2015
17
The findings show that expressions of negative emotions such as anger, stress, and fatigue in the tweets
from people in a given county were associated with higher heart disease risk in that county.
On the other hand, expressions of positive emotions like excitement and optimism were associated with
lower risk.
The results suggest that using Twitter as a window into a community’s collective mental state may provide a
useful tool in epidemiology…So predictions from Twitter can actually be more accurate than using a set of
traditional variables.
18. “With the addition of NLG [Natural Language Generation], smart data
discovery platforms automatically present a written or spoken context-based
narrative of findings in the data that, alongside the visualization, inform the
user about what is most important for them to act on in the data.”
Gartner, 29 June, 2015
Smart Data Discovery Will Enable
New Class of Citizen Data Scientist
19. Systems of Insight
Automated pattern extraction
Outlier detection
Correlation
Time series
Analytics integration with process, app or IoT
https://ubereats.com/melbourne/
21. Reports
&
Analysis
Visualisation
&
Interpretation
Write
Data/Business
“Story”
Insights
Led by Data Analyst or
Scientist
SME owner or Corporate , Machine Learning and Natural Language Generation
Fusion of data science, business knowledge & creativity for maximium ROI
Data
Aggregation Operationalise
Detect &
Extract
Patterns and
Relationships
Generate
Insights &
Story
Process
Application
IoT
Data
Aggregation
or
Data Set
Traditional Analytics: Slow & Expensive
80% of time sifting through data
System of Insight (SoI)
SoI: Fast & Cost Effective
80% of time in decision making with client
22. 22
outlier-detection “allow detecting a significant
fraction of fraudulent cases…different in nature from
historical fraud…resulting in a novel fraud pattern”
Baesens, B., Vlasselaer, V., and Verbeke, W., 2015, Fraud Analytics Using Descriptive,
Predictive, and Social Network Techniques: A Guide to Data Science for Fraud
Detection, Wiley
23. Online tenure leads to more spending per customer
High engagement leads to more orders, more
categories purchased, and more spend
https://www.quillengage.com
24. Better customer experiences . . .
. . . and half the inventory-carrying
costs
of other online fashion retailers.
Forrester, 2016
25. The ANZ Heavy Traffic Index comprises
flows of vehicles weighing more than 3.5
tonnes (primarily trucks) on 11 selected
roads around NZ. It is contemporaneous
with GDP growth.
The ANZ Light Traffic Index is made up of
light or total traffic flows (primarily cars and
vans) on 10 selected roads around the
country. It gives a six month lead on GDP
growth in normal circumstances (but
cannot predict sudden adverse events such
as the Global Financial Crisis).
http://www.a http://www.anz.co.nz/about-us/economic-markets-research/truckometer/
ANZ TRUCKOMETER
26. Systems of Insight
• Helps move away from “crisis levels” in talent
• Traditional 5 step analytics process reduced to 2 step from data to action
• Reimagine business processes through “machine engineering”
• Minimise messy data issues and data preparation time
30. 30
The future is impossible to predict.
However one thing is certain :
The company that can excite it’s customers dreams
Is out ahead in the race to business success
Selling Dreams, Gian Luigi Longinotti