The Degree of Anonymity

We are often faced with the challenge of transparency. How transparent should we be on the internet to ensure we don’t give away too much information and still, get that free account on Spotify or a date on Tinder?

In the article Social Networking Virtues, Vallor points out that anonymity has low entry and exit barriers as compared to social environments (online and offline) but is heavily correlated with less result. So, we can see a linear relationship between transparency and data concern combined with what you’re trying to achieve from the internet be it ordering a perfume for your friend or surfing through HBO. But what is the optimal point? Maybe it varies from site to site or maybe on every individual person based on how worried they are about their data privacy.

Transparency vs Data Concern and Result

Netflix unintentionally stumbled upon this topic back in 2006 when it rolled out a $1M prize for a contest that challenged entrants to improve its recommendation algorithm by 10%. The Netflix Prize was an open competition for the best collaborative filtering algorithm to predict user ratings for films, based on the data Netflix collected from its users. Netflix provided a data set of over 17,000 movies generated by 480,0189 anonymous users without any other information about the users or films. It was a win-win situation: Netflix received relatively cheap R&D, while statistics enthusiasts had a shot at winning a big payday.

Winners of Netflix Prize

Although the data sets were constructed to preserve customer privacy, the Prize has been criticized by privacy claiming that Netflix gave away private information. In 2007, two researchers from The University of Texas at Austin were able de-anonymize user’s identity if that user had also left movie ratings at another site, such as IMDB. Companies like Netflix, Facebook, Google, and Apple are always releasing anonymous data to the public to allow them to build upon the data and in return improve their software and service. I’m sure they want to keep this data as anonymous as they can to avoid the economic ramifications and still get the desired result. But they have no idea of the optimal point is between transparency and result. Does anyone know how anonymous we should be on the internet? The irony is that it wouldn’t be long till a startup becomes a tech giant by guiding people on how transparent they should be and unintentionally ends up using that data set.