COMP 381
Exercise: Data Collection 1.Who are the “fact collectors”? (make a list—be specific) 2.What KINDS of ‘facts’ are stored about you/us? (make a list—be specific)
Fact Collectors (class list) government census taxes work cars, houses, real estate voting record and party affiliation bank accounts medical records criminal history; civil suits guns birth/death insurance schools grades finances address family background SSN medical/mental health salary degrees/education schedule on-campus purchases access to buildings/dorms computer usage/printing activities criminal records employers conflict of interest investments political activities knowledge websites phone calls purchases SSN companies that I buy from buy sites credit ratings address & phone other companies purchased info from other companies info based on IP address public info (see above) Google family and friends
Data & informational privacy One should ask: Who has ACCESS? Who should have ACCESS? Need to know? Why? How long is the data kept?
Question Is there a problem with a search engine monopoly (or oligolopy)? Mowshowitz and Kumar, And Then There Were Three Mowshowitz and Kumar, And Then There Were Three
Search Engine US Market Shares (December 2010) Google 66.6 Yahoo 16.0 Microsoft12.0 Ask 3.5 AOL 1.9 Source: Global
Video Entertainment No Place to Hide No Place to Hide Big Brother Big Business Big Brother Big Business
Data Mining Biometrics I believe you have the right to privacy, but not the right to anonymity -No Place To Hide The laws haven’t caught up to the technology -Big Brother Big Business
Grocery Store Receipts eggs milk bread cheese plates napkins trash bags eggs cheese sour cream bread chips soup milk eggs bread butter cheese gum soda bread eggs milk cheese diapers juice Diapers and beer? Super Bowl Sunday?
Clique = a tightly-knit group of people Clique = a set of vertices in a graph that are all connected to each other by edges of the graph Maximal Clique = a clique that is not a subset of a larger clique Friend Grouper: Definitions Alice Bob Eva David Carol
online social network is rarely a perfect representation of the real social network new member “acquaintances” or “friends” Friend Grouper: Problem Addressed Real LifeOnline AliceBob Eva David Carol Alice Bob Eva David Carol
Triangle Closure Nodes = People Edges = Social Relationships Individuals with common friends are more likely to become friends Leads to “people you may know”
BabyNames BabyNames Pandora Pandora Half.com Half.com Innocuous
Recommending Movies Predicting Ratings Data-Mining Contest: Release rating data, sans “identifying information” Prize awarded Additional contests: canceled ○ Borderline: Netflix
“people you may know” Ex-boyfriend/girlfriend you don’t want to talk to…EVER Person who used to bully you “reconnect” Person in coma in Australia Borderline: Facebook
Types of Invasions Individual Cameras Governmental Snuggly bear on warrantless wiretapping Snuggly bear Institutional Privacy and social networks