The reading notes record thoughts from things I read. 這網誌是我的一些閱讀後的思考和摘要記錄。My website 我的網頁:

Tracking phobia

The W3C Tracking Protection Working Group is still debating on the do-not-track saga, which I mentioned in a blog article earlier.  They have been on the negotiation table for two years with results nowhere in sight.  Online advertisers advocate control on the use of tracked data.  Privacy advocates want no track at all.  On the latest round of talks, some amazing arguments were put forward.  They were reported at IT World.  Take a look below and make your own judgment.  Are they paranoid or are there real danger?

In the seemingly advertising versus privacy battle, the battle front now moves beyond advertising.  It is about creating profiles of your behavior online and designing algorithms to make decisions about you based on that behavior. It is about gathering that data, mining it, and selling it.

1. The prices you pay

Every Web site you visit knows what operating system you are using.  It receives that information automatically, along with other data that can be used to uniquely identify your machine, such as your IP address.  Together with the profile data obtained from behavioral tracking cookies, the information can be used to decide what prices to show you at online purchase based on your presumed wealth.  The Wall Street Journal reported that Mac users who booked hotel rooms paid an average of US$20 to $30 a night more than those who visited the site using a Windows PC. They were not charged more for the same rooms, but they were shown more expensive rooms which they then booked.

2. You are more than the sum of your Likes

Using just Facebook Likes, or some other apps on opinion or mood, researchers are able to determine with impressive accuracy a person’s gender, age, race, sexual tendencies, political leanings, drug habits, and more.  Some of the correlations were quite unexpected.  This is what data mining is about, making correlations between two or more seemingly unconnected bits of information.  It is also what the enforcement agencies did to track potential criminals and terrorists before they strike.  Such information are now available from data brokers who inferred those conclusions based on other things you have done online.

3. Web-lining

If you apply for a job, a loan, insurance, or many other things recently, you probably started by going online and filling out a form. The company could easily have access to one of your profiles.  If you have been tagged by some data mining researches as having some special characteristics, the employer or the bank would know and would use them in their consideration.  Some online banks are already starting to use Facebook profiles as a way to identify poor credit risks.  In fact, refusing to sell insurance or offer loans to people of certain unfavourable background is illegal.  It is called Red-lining.  On the web, it is called Web-lining.  However, in many scenarios, it is not illegal.  This is an area which has attracted some attention.  Some companies have agreed to not collect or use tracking information for the purpose of making an adverse determination of a consumer’s eligibility for employment, credit standing, health care treatment and insurance underwriting. But these principles are voluntary and not widely adopted.

4. Two wrongs could still undo your rights

Online data profiles are often inaccurate.  Multiple people may use the same browser, confusing the profilers. Or someone could visit web sites at random to pollute their data.  But it does not matter.  You might still be offered unexpected discounts or needlessly penalized by an insurance company based on inaccurate information.  Your incorrect profile data could still be used to make decisions about you in invisible ways.