« The "Duplication of Legal Publications" Issue: Recalling a "forgotten moment in the history of law librarianship in which a prominent law librarian provided leadership on a matter of concern throughout the legal profession." | Main | Keynote Sessions from LITA's 2012 National Forum »
October 9, 2012
Some Thoughts (And A Few Personal Disclosures) On Web Privacy
I’m always fascinated by news concerning online privacy, especially in the context of marketing. I think it’s pretty well established that free services from Google, Facebook, and others are not really free. Our payment is the information we voluntarily provide combined with monitoring our activities on the Internet. A story in CIO highlights the lack of transparency on the part of the largest Internet companies on how they create and update user profiles. It’s not merely a matter of filling in the blanks on a profile page. That’s too easy and too obvious. I know that search habits and clicks form the basis of the implied interests, but how does that really work? The more important questions may be how can we edit that information and control its use.
All we see are the end results of that inferential profile. I use Gmail and I know I get targeted ads based on the content of mail sent to and from my account. Google does have a link “Why this ad?” The short description says it is based on emails from my inbox and offers links where I may manage or opt out of ads. There is another option called “Ads on the web” where I can view some of the interest categories associated with my profile.
They are not all entirely accurate. One shows I'm interested in “Arts & Entertainment - Music & Audio - Urban & Hip-Hop - Rap & Hip-Hop.” I do have an interest in music but it does not extend to the listed genres. Another category is “Books & Literature - Children's Literature.” I admit a fondness to Scooby-Doo videos, though I can’t believe that outweighs the subject searches I perform as a reference librarian. I assume Google hasn’t found a way to turn extensive searches on competition, marketing, and antitrust into an ad bonanza. Google seems to think I like basketball. I guess all those searches for news on the Chicago Blackhawks do not register with our mechanized overlords.
The point of this outpouring of detail is to illustrate Google, Facebook, and others pay detailed attention to what we do. We’ve always known that. What I don’t know is how Google selected basketball over hockey. They are not anything alike outside of they are both team sports and typically played in dual use arenas. It also illustrates how profoundly wrong some of algorithms may be in generating our preferred interests. Some of these inferences may have consequences beyond ads.
The CIO article suggests our profiles contain our political affiliations. That may not matter much to some in the United States. It may make a difference, as CIO notes, for citizens in other countries where politics and violence are heavily associated. Google doesn’t list this in my limited demographic listing it displays. For some reason, however, Google suggests following the Obama campaign whenever I log into Google+. I do read a lot of political news though I find it disturbing that Google is predicting an assumed voting preference on my part based on my web habits. I rarely visit campaign sites.
One suggestion in the article is that the only way to make this process more transparent to the end user is through regulation. I think it is a great idea though I think it would be hard to implement. Congress isn’t known for productivity these days. Moreover, there’s money in this, likely generating hard resistance to changing the privacy landscape. CNN reports on how much an individual is worth to Google and others. The amounts change based on, you guessed it, demographics but in 2010 an individual was worth about $14.70 per thousand searches. There are estimates for Facebook as well.
I wouldn’t suggest that individuals give up Google and the rest. I do suggest, however, that anyone using these services should pay attention to the details they have on us when possible. It’s not only marketing. Who knows, maybe these profiles could become a component of things like credit scores. It seems unlikely, but then again, social security numbers were never meant to be unique personal identifiers. Look how that turned out. [MG]