Top Banner
LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU
12

LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

Dec 14, 2015

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

LOCATING EVERYTHING DIGITALMICHAEL R. EVANS

CHINTAN PATEL

UNIVERSITY OF MINNESOTA

{MEVANS,CPATEL}@CS.UMN.EDU

Page 2: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

VISION<location/>

Page 3: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

ENHANCE PROSPERITY

Commerce: Hyper-local advertising

• Out-“local” Groupon• New York Times, “In Groupon’s $6 Billion Wake, a Fleet of Start-Ups”

• Enhanced location-based services• More targeted ads, more customer interest

Government: Laws, Regulations and Revenues

• Online sales tax• California Law ~= $300 million a year!1

Commercial Applications (Geo-DRM)

• Geo-Digital Rights Management2

• MLB.com, Hulu

1http://articles.latimes.com/2011/jun/30/business/la-fi-amazon-tax-201106302Bishr, M., Wytzisk, A., Morales, J., (2007) “GeoDRM: Towards Digital Management of Intellectual Property Rights for Spatial Data Infrastructures” in Research and Theory in Advancing Spatial Data Infrastructure Concepts, ed. Onsrud, H, Redlands, CA: ESRI Press

Page 4: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

ENHANCE SECURITYGeo-location authentication to strengthen passwords

• Access Restrictions• Limit access based on geo-location

• Applications• Prevent phishing, cyber-crime, identity theft• Internet fraud reduction

• Banks look at location of Internet clients• May have prevented Wikileaks

Geographic restrictions on Internet routes

• Cloud Computing• Geographic restriction of data storage

location and transfer routes• Some countries demand readability of all

data packets passing through territory

Page 5: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

ENHANCE CIVILITY

Anonymity often leads to increased aggression1

• Facilitates cyber-crime, such as phishing attacks, flash mobs• “Flash Mob Crime surges: Social Media an enabler?”

• http://www.cbsnews.com/8301-504943_162-20096316-10391715.html

Knowing location of message sources may…

• Reduce phishing attacks• May reduce incivility• Need to balance with censorship risks

Eric Schmidt of Google

• “True anonymity is too dangerous…”2

1Wikipedia, “Social psychology (psychology) ,” 2011.2http://articles.sfgate.com/2011-08-12/opinion/29879141_1_online-anonymity-collateral-damage-randi-zuckerberg

Page 6: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

CHALLENGE

Page 7: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

HOW HARD IS THIS PROBLEM?

Possibly Undecidable• Fully automatic solutions may not

be possible• May need a collection of humans

and computers

Volunteer Geographic Information• OpenStreetMaps during Haiti

earthquake 2009• DARPA Red Balloon Challenge

AI & Machine Learning• IBM Watson hypotheses generation

Validating geo-location once we have it?

• Authentication & Authorization

An undecidable problem is a decision problem for which it is impossible to construct a single algorithm that always leads to a correct yes-or-no answer.

Page 8: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

VOLUNTEERED GEOGRAPHIC INFORMATION (VGI)

Human geo-locating of media:

• Ability to quickly match objects even at different angles• Can handle changes in seasons

Proven effective for mapping, crisis support, etc

Humans can’t search massive data

• Exhaustion, attention-span, etc• Humans can slowly find exact matches• Computers can quickly find rough matches

Page 9: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

MACHINE LEARNING

IBM’s Watson can play Jeopardy, what else can it play?

• Using clues to perform vast searches, evaluate hundreds of conjectures

James Hays, Alexei A. Efros. IM2GPS: estimating geographic information from a single image. Proceedings of the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2008.

Page 10: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

HOW TO AUTHENTICATE GEO-LOCATION?

Challenge Protocols

• Varying power levels, location-specific questions, etc

Authentic and Secure Geo-Location Reporting

• A RSA keyfob with GPS built in?

Geo-enable Internet Infrastructure

• Geo-fencing, routing restrictions, expiration policies• Geo-hardware??

RSA + GPS

Page 11: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

SOCIAL CHALLENGES

Page 12: LOCATING EVERYTHING DIGITAL MICHAEL R. EVANS CHINTAN PATEL UNIVERSITY OF MINNESOTA {MEVANS,CPATEL}@CS.UMN.EDU.

QUESTIONS?

This work supported by the NSF and the USDOD.