Transcript Data Mining 2003 - Carnegie Mellon University

eCommerce Technology
20-751
Data Mining and
Privacy Technology
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Coping with Information
• Computerization of daily life produces data
– Point-of-sale, Internet shopping (& browsing), credit cards,
banks . . .
– Info on credit cards, purchase patterns, product preferences,
payment history, sites visited . . .
• Travel. One trip by one person generates info on
destination, airline preferences, seat selection, hotel,
rental car, name, address, restaurant choices . . .
• Data cannot be processed or even inspected
manually
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Data Overload
• Only a small portion of data collected is analyzed
(estimate: 5%)
• Vast quantities of data are collected and stored out of
fear that important info will be missed
• Data volume grows so fast that old data is never
analyzed
• Database systems do not support queries like
– “Who is likely to buy product X”
– “List all reports of problems similar to this one”
– “Flag all fraudulent transactions”
• But these may be the most important questions!
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Data Mining
“The key in business is to know something that
nobody else knows.”
— Aristotle Onassis
“To understand is to perceive patterns.”
PHOTO: LUCINDA DOUGLAS-MENZIES
PHOTO: HULTON-DEUTSCH COLL
— Sir Isaiah Berlin
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Data Mining
• Extracting previously unknown relationships from
large datasets
– summarize large data sets
– discover trends, relationships, dependencies
– make predictions
• Differs from traditional statistics
– Huge, multidimensional datasets
– High proportion of missing/erroneous data
– Sampling unimportant; work with whole population
• Sometimes called
– KDD (Knowledge Discovery in Databases)
– OLAP (Online Analytical Processing)
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Taxonomy of Data Mining Methods
Data Mining Methods
Predictive
Modeling
• Decision Trees
• Neural Networks
• Naive Bayesian
• Branching criteria
Database
Segmentation
Link
Analysis
Text
Mining
Deviation
Detection
Semantic Maps
• Clustering
• K-Means
Rule Associa tion
Visualization
SOURCE: WELGE & REINCKE, NCSA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Predictive Modeling
• Objective: use data about the past to predict future
behavior
• Sample problems:
– Will this (new) customer pay his bill on time? (classification)
– What will the Dow-Jones Industrial Average be on October
15? (prediction)
• Technique: supervised learning
– decision trees
– neural networks
– naive Bayesian
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Predictive Modeling
Honest
Tridas
Vickie
Mike
Wally
Waldo
Barney
Crooked
Which characteristics distinguish the two groups?
SOURCE: WELGE & REINCKE, NCSA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Learned Rules in Predictive Modeling
Tridas
Vickie
Mike
Honest = has round eyes and a smile
SOURCE: WELGE & REINCKE, NCSA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Rule Induction Example
Data:
height
short
tall
tall
short
tall
tall
tall
short
hair
blond
blond
red
dark
dark
blond
dark
blond
eyes
blue
brown
blue
blue
blue
blue
brown
brown
class
A
B
A
B
B
A
B
B
Devise a predictive rule to classify a new person as A or B
SOURCE: WELGE & REINCKE, NCSA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Build a Decision Tree
hair
dark
blond
red
short, blue = B
tall, blue = B
tall, brown= B
{tall, blue = A}
short, blue = A
tall, brown = B
tall, blue = A
short, brown = B
Does not completely classify
blonde-haired people.
More work is required
Completely classifies dark-haired
and red-haired people
SOURCE: WELGE & REINCKE, NCSA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Build a Decision Tree
hair
dark
blond
red
short, blue = B
tall, blue = B
tall, brown= B
{tall, blue = A}
Decision tree is complete because
1. All 8 cases appear at nodes
2. At each node, all cases are in
the same class (A or B)
short, blue = A
tall, brown = B
tall, blue = A
short, brown = B
eye
blue
short = A
tall = A
brown
tall = B
short = B
SOURCE: WELGE & REINCKE, NCSA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Learned Predictive Rules
hair
dark
blond
red
B
A
eyes
blue
A
brown
B
SOURCE: WELGE & REINCKE, NCSA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Decision Trees
• Good news: a decision tree can always be built from training
data
• Any variable can be used at any level of the tree
• Bad news: every data point may wind up at a leaf (tree has not
compressed the data)
height
tall
short
eyes
blue
brown
B
hair
blonde
A
brown
B
dark
eyes
hair
blonde
B
B
blue
dark
red
A
B
8 cases, 7 nodes. This tree has not summarized the data effectively
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Database Segmentation (Clustering)
• “The art of finding groups in data” Kaufman &
Rousseeuw
• Objective: gather items from a database into sets
according to (unknown) common characteristics
• Much more difficult than classification since the
classes are not known in advance (no training)
• Examples:
– Demographic patterns
– Topic detection (words about the topic often occur together)
• Technique: unsupervised learning
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Clustering Example
• Are there natural clusters in the data (36,10), (12,8),
(38,42), (13,6), (36,38), (16,9), (40,36), (35,19),
(37,7), (39,8)?
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Clustering
• K-means algorithm
• To divide a set into K clusters
• Pick K points at random. Use them to divide the set
into K clusters based on nearest distance
• Loop:
– Find the mean of each cluster. Move the point there.
– Redefine the clusters.
– If no point changes cluster, done
• Clustering demo
• Agglomerative clustering: start with N clusters & merge
• Agglomerative clustering demo
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Neural Networks
Networks of processing units called neurons. This is the j th neuron:
Neuron computes a linear
function of the inputs
n INPUTS
x1, …, xn
1 OUTPUT yj
depends only on
the linear function
Neurons are
easy to simulate
n WEIGHTS
w1j , …, wnj
SOURCE: CONSTRUCTING INTELLIGENT AGENTS WITH JAVA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Neural Networks
INPUT
LAYER
HIDDEN
LAYER
INPUTS:
1 PER INPUT
LAYER NEURON
OUTPUT
LAYER
OUTPUTS:
1 PER OUTPUT
LAYER NEURON
DISTINGUISHED OUTPUT
(THE “ANSWER”)
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Neural Networks
Learning through back-propagation
1. Network is trained by giving it many inputs whose output is known
2. Deviation is “fed back” to the neurons to adjust their weights
3. Network is then ready for live data
DEVIATION
SOURCE: CONSTRUCTING INTELLIGENT AGENTS WITH JAVA
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Neural Network Classification
“Which factors determine a pet’s favorite food?”
Species = Dog
food: Chum
Breed = Mixed
food: Mr. Dog
Owner’s age > 45
Owner’s sex = F
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Neural Network Demos
• Demo: Notre Dame football, Handwriting analyzer
• Financial applications:
– Churning: are trades being instituted just to generate
commissions?
– Fraud detection in credit card transactions
– Kiting: isolate float on uncollected funds
– Money Laundering: detect suspicious money transactions
(US Treasury's Financial Crimes Enforcement Network)
• Insurance applications:
– Auto Insurance: detect a group of people who stage
accidents to collect on insurance
– Medical Insurance: detect professional patients and ring of
doctors and ring of references
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Rule Association
• Try to find rules of the form
IF <left-hand-side> THEN <right-hand-side>
• (This is the reverse of a rule-based agent, where the rules are
given and the agent must act. Here the actions are given and
we have to discover the rules!)
• Prevalence = probability that LHS and RHS occur
together (sometimes called “support factor,” “leverage” or “lift”)
• Predictability = probability of RHS given LHS
(sometimes called “confidence” or “strength”)
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Association Rules from
Market Basket Analysis
• <Dairy-Milk-Refrigerated>  <Soft Drinks Carbonated>
– prevalence = 4.99%, predictability = 22.89%
• <Dry Dinners - Pasta>  <Soup-Canned>
– prevalence = 0.94%, predictability = 28.14%
• <Paper Towels - Jumbo>  <Toilet Tissue>
– prevalence = 2.11%, predictability = 38.22%
• <Dry Dinners - Pasta>  <Cereal - Ready to Eat>
– prevalence = 1.36%, predictability = 41.02%
• <American Cheese Slices >  <Cereal - Ready to Eat>
– prevalence = 1.16%, predictability = 38.01%
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Use of Rule Associations
• Coupons, discounts
– Don’t give discounts on 2 items that are frequently bought
together. Use the discount on 1 to “pull” the other
• Product placement
– Offer correlated products to the customer at the same time.
Increases sales
• Timing of cross-marketing
– Send camcorder offer to VCR purchasers 2-3 months after
VCR purchase
• Discovery of patterns
– People who bought X, Y and Z (but not any pair) bought W
over half the time
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Finding Rule Associations
• Example: grocery shopping
• For each item, count # of occurrences (say out of 100,000)
apples 1891, caviar 3, ice cream 1088, pet food 2451, …
• Drop the ones that are below a minimum support level
apples 1891, ice cream 1088, pet food 2451, …
• Make a table of each item against each other item:
apples ice cream pet food
apples
1891
685
24
ice cream
-----
1088
322
pet food
-----
-----
2451
• Discard cells below support threshold. Now make a cube for
triples, etc. Add 1 dimension for each product on LHS.
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Rule Association Demos
• Magnum Opus (RuleQuest, free download)
• See5/C5.0 (RuleQuest, free download)
• Cubist numerical rule finder (RuleQuest, free
download)
• IBM Interactive Miner
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Text Mining
• Objective: discover relationships among people &
things from their appearance in text
• Topic detection, term detection
– When has a new term been seen that is worth recording?
• Generation of “knowledge map”, a graph
representing terms/topics and their relationships
• SemioMap demo (Semio Corp.)
–
–
–
–
Phrase extraction
Concept clustering (through co-occurrence) not by document
Graphic navigation (link means concepts co-occur)
Processing time: 90 minutes per gigabyte
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Major Data Mining Ideas
•
•
•
•
There’s too much data
We don’t understand what it means
It can be handled without human intervention
Relationships can be discovered automatically
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Privacy Policies
• Privacy policy:
– what data we collect
– what we do with the data
– how we control its release
• A privacy policy is written in natural language, e.g.
English
• How can a machine understand it?
• Answer: express the policy in a formal, machinereadable way
• P3P for website policies
• EPAL for enterprise policies
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
P3P
• Platform for Privacy Preferences
• Idea: allow the browser to control the release of data
to a website
• Website states its policy in machine-readable form
(P3P)
• User expresses his privacy preferences in machinereadable form
• Browser compares Website policy with user
preferences and controls release of data
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Simple HTTP Transaction
Web
Server
GET /index.html HTTP/1.1
Host: www.att.com
. . . Request web page
HTTP/1.1 200 OK
Content-Type: text/html
. . . Send web page
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
SOURCE: LORRIE CRANOR
HTTP With P3P 1.0
GET /w3c/p3p.xml HTTP/1.1
Host: www.att.com
Web
Server
Request Policy Reference File
Send Policy Reference File
Request P3P Policy
Send P3P Policy
GET /index.html HTTP/1.1
Host: www.att.com
. . . Request web page
HTTP/1.1 200 OK
Content-Type: text/html
. . . Send web page
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
SOURCE: LORRIE CRANOR
Assertions in a P3P policy
• General assertions
– Location of human-readable policies and opt-out mechanisms –
discuri, opturi attributes of <POLICY>
– Indication that policy is for testing only – <TEST> (optional)
– Web site contact information – <ENTITY>
– Access information – <ACCESS>
– Information about dispute resolution – <DISPUTES> (optional)
• Data-Specific Assertions
– Consequence of providing data – <CONSEQUENCE> (optional)
– Indication that no identifiable data is collected –
<NON-IDENTIFIABLE> (optional)
– How data will be used – <PURPOSE>
– With whom data may be shared – <RECIPIENT>
– Whether opt-in and/or opt-out is available – required attribute of
<PURPOSE> and <RECIPIENT>
– Data retention policy – <RETENTION>
– What kind of data is collected – <DATA>
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
SOURCE: LORRIE CRANOR
P3P/XML encoding
Statement
P3P version
<POLICIES xmlns="http://www.w3.org/2002/01/P3Pv1">
<POLICY discuri="http://p3pbook.com/privacy.html"
Location of
name="policy">
human-readable
P3P policy name
<ENTITY>
<DATA-GROUP>
privacy policy
<DATA
Site’s
ref="#business.contact-info.online.email">[email protected]
name
</DATA>
and
<DATA
ref="#business.contact-info.online.uri">http://p3pbook.com/
contact
</DATA>
info
<DATA ref="#business.name">Web Privacy With P3P</DATA>
</DATA-GROUP>
Access disclosure
</ENTITY>
Human-readable
<ACCESS><nonident/></ACCESS>
explanation
<STATEMENT>
<CONSEQUENCE>We keep standard web server logs.</CONSEQUENCE>
<PURPOSE><admin/><current/><develop/></PURPOSE>
How data may
<RECIPIENT><ours/></RECIPIENT>
be used
<RETENTION><indefinitely/></RETENTION>
<DATA-GROUP>
Data recipients
<DATA ref="#dynamic.clickstream"/>
<DATA ref="#dynamic.http"/>
Data retention policy
</DATA-GROUP>
</STATEMENT>
Types of data collected
</POLICY>
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
</POLICIES>
SOURCE: LORRIE CRANOR
P3P Increases Privacy Awareness
http://www.att.com/accessatt/
• P3P clients can check a
privacy policy each time
it changes
• P3P clients can check
privacy policies on all
objects in a web page,
including ads and
invisible images
http://adforce.imgis.com/?adlink|2|68523|1|146|ADFORCE
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
SOURCE: LORRIE CRANOR
P3P in IE6
Automatic processing of
compact policies only;
third-party cookies without
compact policies blocked by
default
Privacy icon on status bar
indicates that a cookie has
been blocked – pop-up appears
the first time the privacy icon
appears
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
SOURCE: LORRIE CRANOR
Users can click on
privacy icon for
list of cookies;
privacy summaries
are available at
sites that are
P3P-enabled
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
SOURCE: LORRIE CRANOR
Typical Anonymizer
Request
Request
Anonymizer
Reply
Reply
Client
Server
•
•
•
•
Acts as a proxy for users
Hides information from end servers
Sees all web traffic
Adds ads to pages (free service; subscription service
also available)
http://www.anonymizer.com
SOURCE: LORRIE CRANOR
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Mixes
Sender
Destination
B, C
dest,msg kC kB kA
Mix C
msg
dest,msg kC
Mix A
C dest,msg kC kB
Mix B
kX = encrypted with public key of Mix X
Sender routes message randomly through network
of “Mixes”, using layered public-key encryption.
SOURCE: LORRIE CRANOR
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Crowds
• Users join a Crowd of other users
• Web requests from the crowd cannot be linked to any
individual
• Protection from
–
–
–
–
end servers
other crowd members
system administrators
eavesdroppers
• First system to hide data shadow on the web without
trusting a central authority
http://www.research.att.com/projects/crowds/
SOURCE: LORRIE CRANOR
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
EPAL
• P3P is designed for web privacy
• EPAL describes internal privacy policies
• Data Users: Individuals who are accessing or receiving data.
Examples: nurse, doctor
• Actions: Examples: create a customer record, read a record.
• Data Categories: High-level descriptions of types of data.
Examples: customer contact information, medical information
• Purposes: How the data will be used by the recipient.
Examples: marketing, fulfillment.
• Conditions: Examples: age < 13, consent obtained.
• Obligations: Additional steps to be taken when access occurs.
Examples: log the access, delete data after 1 year if no
additional business
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
EPAL
• ALLOW [Data User] TO PERFORM [Action] ON [Data Type]
FOR [Purpose] IF [Condition] AND CARRY OUT [Obligation]
• DENY [Data User] TO PERFORM [Action] ON [Data Type] FOR
[Purpose] IF [Condition] AND CARRY OUT [Obligation]
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
An EPAL Rule
• English: “Email can be used for the book-of-month club only if
consent has been given and age is more than 13.”
<rule id="rule1" ruling="allow">
<data-user id="borderless-books"/>
<data-category id="email"/>
<purpose id="book-of-the-month-club"/>
<action id="read"/>
<condition id="consentToBookClub"/>
<condition id="olderThan13"/>
<obligation id="retention">
<parameter id="days">5</parameter>
</obligation>
</rule>
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
EPAL
SOURCE: IBM, W3C
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS
Q&A
20-751 ECOMMERCE TECHNOLOGY
SUMMER 2003
COPYRIGHT © 2003 MICHAEL I. SHAMOS