Data Science in Context

Privacy & Ethics in Data Mining

Collecting Detailed Consumer Information

Businesses rely on consumer data to understand their customers and make predictions.

  1. There is a market for customer information.
  2. Access to this data grants businesses the ability to:
    • Target marketing campaigns
    • Predict Trends
  3. Privacy regulations affect the availability of some of this data (i.e. location)
  4. Researchers have identified direct links to privacy restrictions and business performance.

From the text...

Case Study "Privacy Regulation and Online Advertising"
Who?: MIT and UT Researchers: A. Goldfarb and C. Tucker
When?: August 5, 2010
What: 9,596 online advertisements; 3.3 million survey-takers
How Cross-section analysis of advertisement: purchase intent, ad recall, & product favoribility
Why: Examine the effectiveness of online advertising with increased protection measures

Key take-aways:

  1. Ads are most effective when they appeal to their audience.
  2. Higher levels of loss of effectiveness for:
  3. General content websites (i.e. news)
  4. Non-interactive ads (i.e. plain banner)
  5. Laws reduced the effectiveness of advertising by over 65%, but it is 65% of a relatively small number. (1.71% decrease in effectiveness)

EU General Data Protection Regulation (GDPR)

Purpose: "To give citizens back control over their personal data."

  1. Supercedes Data Protection Directive
  2. Four years of negotiation; Effective May 25, 2018.
  3. Established privacy goals for all member countries
  4. Affects any company that used EU citizen data
  5. Restricts transfer of data outside of the EU
  6. Guarentees right to data breach notifications

Personal Data, GDPR definition

"Information relating to an identified or identifiable natural person."

Personal data protections now include information such as:

  • IP address
  • Device ID
  • Customer reference number

Some exceptions for data for scientific and statistical purposes.

GDRP Reception

Pros

  1. Guarenteed protections for users
  2. Increased data hygiene

Cons

  1. Increased legal complexities
  2. Increased bureaucratic burden

Privacy in the News

url <- "https://newsapi.org/v2/everything?q=data-privacy&from=2018-12-04&to=2018-12-04&sortBy=relevancy"
key <- "&apiKey=6b6e9a9d6662400e9d91276433535f55"
news <- fromJSON(paste0(url, key))
Most Relevant News Stories from 12/04/2018
Articles Pertaining to Data Privacy
‘Google You Owe Us’ claimants aren’t giving up on UK Safari workaround suit
Why AOL was just hit with a record fine over privacy concerns
Your Money: Marriott Hack Adds Passport Headache, but Its Intensity Is in Dispute
What's Wrong With Your Venmo, and How to Fix It
Wish You Were Smarter? Check out These 4 Brain Hacks That Can Help You Get There.

U.S. Comparisions

Legislation tends to arise on an ad hoc basis

  • Video Privacy Protection Act of 1988
  • Cable Television Protection and Competition Act of 1992
  • Fair Credit Reporting Act,
  • 1996 Health Insurance Portability and Accountability Ac (HIPAA)

Conclusion

Ethical usage of privacy data requires:

  • Transparency
  • Trust
  • Purpose
  • Compliance

There will be many more futher considerations for ethical data usage as time and technology continue to advance.

Sources