Practical Privacy Pointers For Libraries

1
Practical Privacy PointersFor Libraries

• Karen A. Coombs
• Electronic Services Librarian
• SUNY Cortland
• coombsk_at_cortland.edu
• http//www.librarywebchic.net/presentations/cil200
  5

2
Why libraries should care about privacy?

• Library Bill of Rights ALA Code of Ethics
• Legislation at the state and Federal Level
• User trust
• Users trust that we are protecting their privacy
  even if we are not or cannot.
• User expectations
• Most commercial websites and corporations have
  some sort of privacy policy.
• In a world where libraries are being

3
Why do we need practical privacy pointers?

• Digital libraries can be wonderful supports for
  spying.
• Roy Tennant
• ALAs Guidelines lack a technical perspective
• Privacy is an onion
• Lots of layers

4
(No Transcript)
5
Where to begin

• Decide on a Privacy Strategy
• Assess Where Use Data Can be Collected
• Understand the effects of the policy you put in
  place
• How is the policy going to effect your data
  collection procedures and library systems?
• What technologies are you going to use to execute
  the policy?

6
Decide on a Privacy Strategy

• Collect only what is absolutely necessary.
• Keep only what you're willing to give up. Destroy
  everything else.
• Roy Tennant 7/15/2003 Issue of Library Journal
• Lots of leeway here
• Degree of vigilance and thoroughness varies from
  library to library and what you consider
  personably identifiable information

7
Assessing Where User Data can be collected

• Integrated Library Systems
• Interlibrary Loan Systems
• Web Servers
• Proxy servers
• Public Computers
• E-reference systems

8
What user data are we currently collecting?

• Name
• Address
• ID Number
• Phone
• Email
• IP address

9
What data do we need to collect?

• Daily operations
• Example user data for circulation
• Security
• Example Username and passwords
• Troubleshooting
• Example Server log files and User IDs
• Assessment
• What questions do you want to answer?

10
Data We Had and Dont Need

• Birthdate
• Social Security Number
• Identity Theft Issues
• Do you have all the information for someone to
  complete a credit card application?
• How is user information the library collects used
  in other campus systems?
• Default PINs

11
Thinking About Protecting Privacy

• Implemented ALEPH
• ALEPH designed by Ex Libris
• Not based in United States
• Many initial clients in Europe and other parts of
  the world
• European privacy laws very different from US
• Link between user and item borrowed not broken
  when items returned

12
Integrated Library System

• Older versions of ALEPH do not break the link
  between user and the items they borrowed when
  items are returned
• Data about who borrowed what is stored in three
  tables
• Oracle relentlessly saves data
• Backups
• Log files

13
Integrated Library System

• Ex Libris developed script to de-link user
  information from items
• Alters three tables in ALEPH to remove user id
• Can still get generic demographic information
  about user
• Status (faculty, staff, student)
• Run delinking script before your backups

14
WebOPAC Server Log Files

• IP address
• Can contain users library id
• Search user performed
• Index searched
• Search terms
• Cleaned out weekly

15
Interlibrary Loan System

• ILLiad
• SQL Server database
• Table for Transactions
• Each transaction is associated with a specific
  username
• Keeps track of what users borrow

16
Interlibrary Loan System

• Atlas provides SQL script to break link between
  user and items interlibrary loaned
• If use script lose all demographic information
• Wanted to keep demographic info
• Created new script
• Writes demographic information to a new table
• Breaks link between user and items interlibrary
  loaned
• Need to synch script with SQL server backups

17
Web Servers

• Contain IP address of user
• Collected as daily files
• Last octet of IP address masked at end of day
• Why not mask entire IP address?
• Collect statistics about where users are using
  library resources
• Subnets

18
Mechanics of log washing

• Determine the previous days date
• Wash log files for previous day
• Add information from previous days log file to
  web server statistics

19
EZProxy Server

• Contain IP address of user
• Can contain users ID information
• Prevent user id from being written to proxy
  server log files
• Currently
• Collected as monthly files
• Monthly files discarded after processed for
  statistical information
• Future Plans
• Collect as daily files
• Mask last octet at end of day
• Why not mask entire IP address?
• Collect statistics about where users are using
  library resources
• Subnets

20
Open URL Resolver

• SFX
• Has log files which contain
• Users IP address
• OpenURL request
• Citation for item
• Server Maintained by the SUNY Office of Library
  and Information Services
• Log files dont seem to be purged on a regular
  basis

21
Public Computers

• Configuration
• Set Temporary Internet Files to 2MB
• Set History to zero days
• Leave Cookies on
• Form Memory turned off
• Deep Freeze
• Clears computers when they are rebooted
• Fall 2005
• All computers will require authentication
• Pros
• Cache will be cleared when user logs out
• If computers are left logged in and inactive they
  will rebooted
• Cons
• Non-college users will have to obtain a guest
  user account to use library computers

22
Lessons Learned

• A Privacy Policy is more than a written document
• You dont always control the systems where your
  user data is being stored
• Data can be stored in places you didnt think of
• How you purge user data can effect your libraries
  ability to function

23

• Questions?
• Further Resources
• ALA Guidelines for Developing a Library Privacy
  Policy
• Web Log Washing
• http//sunsite.berkeley.edu/Web4Lib/RefCenter/logw
  ashing.html