and in the air Comparing Perches and Perspectives SSLBC2012 Whats Up 2 Why am I here Everyone else was busy Excuse to visit campus Wouldnt have finished my thesis if it werent for Ellen ID: 591275
Download Presentation The PPT/PDF document "Elections on the ground" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Elections on the ground and in the air
Comparing Perches and Perspectives
#SSLBC2012Slide2
What’s Up?
2Slide3
Why am I here?
Everyone else was busyExcuse to visit campus
Wouldn’t have finished my thesis if it weren’t for Ellen
Follow and have worked in political risk
Helped managed election observation mission in Liberia, November 2011
Currently, working at startup that uses predictive analysis and temporal analytics
3Slide4
Fields of Reference
4Slide5
Outline
Definitions, Perspectives, and Engaging
Futures & Forecasting Methodologies
Case Study: Egypt
Online Tools: Whither Twitter?
Temporal
Vision
5Slide6
When I use a word it means just what I choose it to mean — neither more nor less
.
Humpty Dumpty in Lewis Carroll’s Through the Looking Glass
1
.
D
efinitions
, Perspectives,
& Engaging ElectionsSlide7
Four Ways to Manage Info
7Slide8
Definitions
Fact: evidence used in reportOpinion: biased perspective
Data: quantity, number, sum
Information: relationships of facts
Intelligence: organized information
8Slide9
Perspective, Parsed
9Slide10
Engaging Elections
10Slide11
Primary Actors & Key Stakeholders
11Slide12
Central to Elections
12Slide13
Characteristics and Categories
13Slide14
It is hard to make predictions, especially about the future.
Yogi Berra
4. Forecasting MethodologiesSlide15
Phillip K. Dick
15Slide16
Providential:
what will be will beConventional: tomorrow will be much like today
Pessimism
: decline from past ‘Golden Age’
Discontinuity
: the future will be nothing like the present
Optimism
: faith in progress / technology cures all
Unknowable
: futile to attempt to go beyond the present
Futurist
: tempered optimism / the future is rich with possibility resulting from human planning and action
Source Prof. Howard F.
Didsbury
Jr
Seven Attitudes Towards Future
16Slide17
Forecasting Animals
17Slide18
Types of Political Risk
18Slide19
How to Make & Falsify Predictions
Probability
Impact
Time Range
19Slide20
If a man will begin with certainties, he shall end in doubts; but if he will be content to begin with doubts he shall end in certainties.
Francis Bacon
2. Case Study: EgyptSlide21
Time
Window Analysis
21Slide22
Profile: Historic Milestones
Egyptian
Gov’t
Ongoing Arrests of MB Members
President Obama’s Cairo Speech
As Opposition to Mubarak, MB Backed by
ElBaradei
MB Threatens Disruption of Elections
22
Past 5 Years
3 Months prior
1 Month following
Run-up
Up to TodaySlide23
Watersheds
pre-Jan25
23
Past 5 Years
3 Months prior to #Jan25
1 Month following #Jan25
Run-up
Up to TodaySlide24
Watersheds
post-Jan25
24
Past 5 Years
3 Months prior to #Jan25
1 Month following #Jan25
Run-up
Up to TodaySlide25
Evolution: Political Landscape
25
Source:
Arabist.netSlide26
Change
in FJP Composition
Creation of FJP & Before Elections
26
After ElectionsSlide27
When
to Watch
27
Mid-April
: Candidacy Registration Period
Mid to Late June
: anticipated electionsSlide28
Comparison
of January 2012
28
Forecast – 6 Weeks Prior
Actual – 4 Weeks During
Revised – 6 Weeks AfterSlide29
Time
-Textured January 2012
29
Forecast – 6 Weeks Prior
Actual – 4 Weeks During
Revised – 6 Weeks AfterSlide30
What
and How to Watch
30Slide31
Once we get out of the '80s, the '90s are gonna
make the '60s look like the '50s.
Dennis Hopper in
Flashback
4. Online Tools: Whither Twitter?Slide32
Seismic Shift in Intelligence
Temporal Indexing of Web Enables
Novel IntelligenceSlide33
Text
Is Loaded with Temporal Signals
“
North Korea
apparently began pursuing a uranium enrichment program in 1996 at the latest
”
...In June, officials said the
network encryption
was operational
“
2012 is the year when
China
will
export
more chemicals according to this source
”
“
Dr Sarkar
says the new facility will
be operational by March 2014...”
Drought and malnutrition hinder next spring’s expansion plans in
Kabul
...
“...opposition organizers plan to meet on Thursday to
protest
...”
Unstructured text has analytic & predictive power.Slide34
Using Big Data for Prediction
34Slide35
New Tools: Electionista
web app for monitoring elections over Twitter across 110 different countries in 58
languages
Tweets are categorized
geographically
and cached
so that users can scroll
back
35Slide36
New Tools: WaPo Modifiable Model
Uses 3 variables
:
% change in GDP per capita from Q1 to Q3
Average Approval in June, according to Gallup
1 incumbent party candidate is sitting president, 0 if not
SOURCES: Seth Hill, postdoctoral associate at Yale University; John Sides, associate professor at George Washington University; Lynn
Vavreck
, associate professor at UCLA; GRAPHIC: Jeremy Bowers, Emily Chow and Ezra Klein - The Washington Post. Published April 24, 2012.
36Slide37
New Tools: Google Portal
37Slide38
New Tools: Google Portal
38Slide39
New Tools: Yahoo Clues
How people search
Find most popular search terms
Compare trends between search terms
39Slide40
New Tools: RF & Temporal Analysis
40
Sentiment analysis
Temporal network analysis
Search through time
Temporal source scoringSlide41
Further Research for Twitter Data
S
entiment
analysis of political
tweets
Automatic
detection of propaganda and
disinformation
Automatic
detection of sock
puppets
Credibility checking
Basic
research on Twitter demographics and automatic profiling of users with regards to demographic
attributes
Basic
research on user participation and self-selection
bias
41Slide42
Discovery consists of seeing what everybody has seen and thinking what nobody has thought.
Albert Szent-Gyorgyi
5. Temporal VisionSlide43
Temporal Vision
43Slide44
Insight from Big Data
Watch (signal)Have expectations about what is being watched (shift)
Know how is obscured (
blindspot
)
Identify out-of-the-ordinary happenings (outlier)
Conceive of fast-moving, far-reaching events (flashpoints)
Be able to correlate them with other interesting observations (pattern)
44Slide45
Thank You!
E:
munish@recordedfuture.com
T: @
whypurifly
@
politicalrisk
L:
linkd.in/munish
45Slide46
Nobody is wrong about the future and everybody is wrong about the past.
Munish Puri