Transcript NSF CDI meeting - University of Illinois at Urbana

Position paper for CyberGIS’12.
The University of Illinois at Urbana-Champaign, IL
August 7, 2012 (Research Symposium) (NCSA Auditorium)
Mapping the Spatial Diffusion of
Innovations in Cyberspace and the
Distortion Effects via Different Social
Media Communication Channels
• Ming-Hsiang (Ming) Tsou [email protected],
Professor (Geography), San Diego State University, PI of “Mapping Ideas” project.
Co-PIs:
Dr. Dipak K Gupta (Political Science), Dr. Jean Marc Gawron (Linguistic),
Dr. Brian Spitzberg (Communication), Dr. Li An (Geography)
San Diego State University, USA
Funded by NSF Cyber-Enabled Discovery and Innovation
(CDI) program. Award # 1028177. (four years 2010-2014)
http://mappingideas.sdsu.edu/
Special thank to our wonderful
graduate students and
undergraduates:
Current (2012):
Ph.D. Students: Ick Hoi (Rick) Kim, Jiue-An (Jay)
Yang, Sarah Wandersee, Ninghua Wang.
Master Students: Daniel Lusher, Anna Nagel,
Nicole Stotz,
Undergraduates: Alejandra Coronado.
2010 - 2011: Kellen Stephens, Ting-Hwan Lee,
Amit Nagesh, Vickie Mellos, Jennifer Smith, Luke
Kemper, James Banker.
DIFFUSION OF INNOVATIONS
THE INNOVATION DIFFUSION MODEL (E. Rogers, 1962):
Diffusion: “a special type of communication concerned with
the spread of messages that are perceived as new ideas” (E.
Rogers, p. 35).
Torsten Hägerstrand, 1953 – Doctoral Dissertation - Culture
Diffusion. 1967 Innovation Diffusion as a Spatial Process.
Public (Mass Media, TV, News)
vs.
Private (Personal Communications)
Can we apply this diffusion theory in
Cyberspace? Example: Occupy Wall Street ?
Source: http://www.economist.com/blogs/democracyinamerica/2011/10/socialmedia-and-wall-street-protests
Mapping Social
Media and “Ideas”
on Digital Earth.
Can we create a “new method” to
understand and analyze social
activities, ideas, and human
communications on our digital Earth?
(revealing the invisible world)?
Similar to the “remote sensing
satellites” or “social radar” to monitor
human activities on our Earth. This
new method will help us to detect:
•
Image created by Ming-Hsiang Tsou (NSF-CDI project)
The location of 1000 web pages with keyword “white
powers” from Yahoo Search Engine on October 15,
2010.
Social Weather Maps
Create “
”?
(monitoring social “temperature” changes, social
“storms”…)
•
Hot spots of ideas, events, or
concepts. (social storms?)
Dynamic changes of social
activities, ideas, events, and
social movements. (change of
social weather maps)
Spatial Web Automatic Reasoning and
Mapping System (SWARMS ) prototype
Information Communication Channels in Cyberspace
• Web Pages ( Semi-Public Information Communication )
• Social Media (Twitter: Semi-Private Information Communication)
Web Pages: Use Web Search
Engines (Google, Yahoo, and Bing) to
retrieve up to 1,000 web pages per
keyword. Then analyze their contents
associated with their ranks and
geolocations.
Social Media (Tweets): Use Twitter
APIs to retrieve tweets based on
Keywords or #Hashtag and
geolocations (self-defined home-towns
or GPS locations).
Collect Web Page Contents, Ranks, and Locations:
We develop Cyber-Discovery Search Engine
(Retrieve up to 1000 results from Yahoo or Bing)
“Burn Koran” Web Information Landscape
The kernel density of
“burn Koran” points (keyword
search results) from 1000
associated websites (red
dots) Yahoo search engine
on January 30, 2011.
The kernel density of
50,000 web pages from
300 random keywords.
(Baseline map)
“burn Koran” of Web Information Landscape
Baseline (average) of Web Information Landscape
Differential of Web Information Landscape
Social Media (Tweets) –Semi-Private Communication:
Use Twitter APIs to retrieve tweets based on Keywords or #Hashtag and geolocations
(self-defined home-towns or GPS locations).
Twitter






An online social networking service
A micro-blogging service
Created in March, 2006
Reached 300 million users as of 2011
Over 300 million tweets per day
Over 1.6 billion search queries per day
Twitter Application Programming Interface (API)



For software to communicate with other programs
Search API: query for Twitter content + spatial query function.
Streaming API: real-time data trend mining
TWITTER – SPATIAL SEARCH API
Center: 41.961295, -93.281859
Radius: 180 miles
Limitations:
1. Spatial Search can only trace back up to seven
days. (Regular search can trace back to 14 days.)
2. Each search results can not exceed 1500
tweets.
Tweets Search Keyword: “Mitt Romney”
Output contents export to Excel files (.xls)
Tweets% (one day before), tweets% (election day),
and the Actual Election Results%
Sometimes the one
day before tweets% is
closer to the actual
election results.
Ron Paul % has very
big “distortion”
comparing to the
actual results.
All cyberspace maps are “distorted from the real
world due to the “projection methods” -communication methods or media characteristics.
Different Topics
Real World
Different Projection
Methods (Distortion)
What kinds of “distortion”?
• Explanation
• Space-Time
Analysis
• Prediction?
Correct the distortion effects
Epilogue: Big Brother (1984)?
What types of ethical and civil-liberty implications (in terms of
privacy, security and human rights) will need to be considered in
the context of Web surveillance technologies and the analytical
tools of web and social media?
Thank You
[email protected]
Funded by NSF Cyber-Enabled Discovery and Innovation (CDI) program.
Award # 1028177. (four years 2010-2014) http://mappingideas.sdsu.edu/