Transcript Against Data-Mining Uses

Against Data-Mining Uses
Team 10
Ben Jabara
Jesse Lentz
Ben Heller
Background
 Uses data warehouses comprised of large
amounts of data to find trends
 Increased usage after terrorist attacks (Dept. of
Homeland Security)
 Data Mining Uses
–
–
–
–
–
improving service or performance
detecting fraud, waste, and abuse
analyzing scientific and research information
detecting criminal activities or patterns
analyzing intelligence and detecting terrorist activities
Sources of Data
 Directly from individuals
– Surveys
 Web Crawlers
– Indexing Services like search engines
 Government Agencies
– National Oceanic and Atmospheric Administration
 Private Sector Sources
– Credit Card Companies
– Banks and other Financial Institutions
GAO Study of Data Mining
 Government agencies required to provide notice to
individual respondents explaining why the information is
being collected
– Only 2 of 5 agencies in GAO study provided this notice
– Agency compliance with key security requirements was
inconsistent
 122 of 199 data mining efforts used personal information
–
–
–
–
student loan application data
bank account numbers
credit card information
taxpayer identification numbers
 Results: “Until agencies fully comply with these
requirements, they lack assurance that individual privacy
rights are being appropriately protected”
Inherent Data Mining Issues
 Can reveal patterns and relationships but
doesn’t tell the value or significance of these
patterns
 Can identify connections between behaviors
and/or variables
 DOES NOT necessarily identify a causal
relationship
– Leads to false positives and incorrect
conclusions
Data Mining Ineffectiveness
The New York Times confirmed that point, with its report that
the NSA has sent the FBI a "flood" of tips generated by
mass domestic eavesdropping and data mining, virtually all
of which led to dead ends that wasted the FBI's resources.
"We'd chase a number, find it's a schoolteacher with no
indication they've ever been involved in international
terrorism," one former FBI agent told the Times. "After you
get a thousand numbers and not one is turning up
anything, you get some frustration.”
ACLU.org - Eavesdropping 101: What Can The NSA Do?
Data Mining Ineffectiveness
1) Detect: Terrorists Enjoy Driving El Caminos
2) Classify: El Camino lovers classifies person as potential
terrorist
3) ID: John Bagby internet queries reveals El Camino
interests
4) Track: Bagby tracked searching for El Camino availability
at www.autotrader.com
5) Understand: Bagby works for a university that does
research for sensitive government projects
6) Preempt: Bagby’s door kicked down by government
agents
Privacy Violations
 Pentagon’s Total Information Awareness
System
– Violation of 4th Amendment rights because
warrants needed to search a particular place as
opposed to blanket searches
– Right of people to be secure…in their papers
 Data mining and warehousing efforts hastily
implemented after 9/11
– Few privacy considerations
Further Problems
 Compromise of personal data by hackers or
data misuse
– Many data warehousing efforts performed by
private businesses