1. Privacy Preserving Data Mining
Ping Chen Lila Ghemri
U of Houston –Downtown Texas Southern University
One Main Street Cleburne St
Houston, Texas Houston, Texas

2. Topics Overview Basic methods in PPDM Association rule mining
Classification

3. Individual Privacy: Protect the “record”
Individual item in database must not be disclosed
Not necessarily a person
– Information about a corporation
– Transaction record
Disclosure of parts of record may be allowed
– Individually identifiable information

4. Individually Identifiable Information
Data that can’t be traced to an individual not
viewed as private
– Remove “identifiers”
But can we ensure it can’t be traced?
– Candidate Key in non-identifier information
– Unique values for some individuals

5. PPDM Methods Data Obfuscation – Nobody sees the real data
Summarization
– Only the needed facts are exposed
Data Separation
– Data remains with trusted parties

6. Data Obfuscation Goal: Hide the protected information Approaches
– Randomly modify data
– Swap values between records
– Controlled modification of data to hide secrets
Problems
– Does it really protect the data?
– Can we learn from the results?

7. Summarization Goal: Make only innocuous summaries of data available
Approaches:
– Overall collection statistics
– Limited query functionality
Problems:
– Can we deduce data from statistics?
– Is the information sufficient?

8. Data Separation Goal: Only trusted parties see the data Approaches:
– Data held by owner/creator
– Limited release to trusted third party
– Operations/analysis performed by trusted party
Problems:
– Will the trusted party be willing to do the analysis?
– Do the analysis results disclose private information?

9. Association Rules Association rules a common data mining task
– Find A, B, C such that AB → C holds frequently
Fast algorithms for centralized and distributed computation
– Basic idea: For AB → C to be frequent, AB, AC, and BC must all be frequent
– Require sharing data
Secure Multiparty Computation too expensive – Given function f and n inputs distributed at n sites, compute f(x1, x2,…,xn) without revealing extra information.

10. Association Rule Mining: Horizontal Partitioning
Distributed Association Rule Mining: Easy without sharing the individual data (Exchanging support counts is enough)
What if we do not want to reveal which rule is supported at which site, the support count of each rule, or database sizes?
• Hospitals want to participate in a medical study
• But rules only occurring at one hospital may be a result of bad practices

11. Association Rules in Horizontally Partitioned Data

12. Overview of the Method
Find the union of the locally large candidate itemsets securely
After the local pruning, compute the globally supported large itemsets securely
At the end check the confidence of the potential rules securely

13. Securely Computing Candidates
Key: Commutative Encryption (E1(E2(x)) = E2(E1(x))
• Compute local candidate set
• Encrypt and send to next site
• Continue until all sites have encrypted all rules
• Eliminate duplicates
• Commutative encryption ensures if rules the same, encrypted rules the same, regardless of order
• Each site decrypts
• After all sites have decrypted, rules left
Care needed to avoid giving away information through ordering/etc.
Redundancy maybe added in order to increase the security.

14. Computing Candidate Sets

15. Compute Which Candidates Are Globally Supported?

16. Which Candidates Are Globally Supported? (Continued)
Now securely compute Sum ≥ 0:
- Site0 generates random number R and
sends R + count0 – frequency * dbsize0 to Site1
- Sitek adds countk – frequency*dbsizek and sends it
to Sitek+1
Final result: Is sum at Siten – R ≥ 0?
Use secure two-party computation

17. Computing Frequent: Is ABC 5%?

18. Computing Confidence

19. Classification by Decision Tree Learning
A classic machine learning / data mining problem
Develop rules for when a transaction belongs to a class based on its attribute values
Smaller decision trees are better
ID3 is one particular algorithm

20. A Database…
Outlook Temp Humidity Wind Play Tennis Sunny Hot High Weak No Sunny Hot High Strong No Overcast Mild High Weak Yes Rain Mild High Weak Yes Rain Cool Normal Weak Yes Rain Cool Normal Strong No Overcast Cool Normal Strong Yes Sunny Mild High Weak No Sunny Cool Normal Weak Yes Rain Mild Normal Weak Yes Sunny Mild Normal Strong Yes Overcast Mild High Strong Yes Overcast Hot Normal Weak Yes Rain Mild High Strong No

21. … and its Decision Tree Outlook Sunny Rain Overcast Humidity Wind Yes
High
Normal
Strong
Weak
Yes No
Yes No

22. The ID3 Algorithm: Definitions
R: The set of attributes
Outlook, Temperature, Humidity, Wind
C: the class attribute
Play Tennis
T: the set of transactions
The 14 database entries

23. The ID3 Algorithm ID3(R,C,T)
If R is empty, return a leaf-node with the most common class value in T
If all transactions in T have the same class value c, return the leaf-node c
Otherwise,
Determine the attribute A that best classifies T
Create a tree node labeled A, recur to compute child trees
edge ai goes to tree ID3(R - {A},C,T(ai))

24. The Best Predicting Attribute
Entropy
Gain(A) =def HC(T) - HC(T|A)
Find A with maximum gain