Fixing Facebook: Attempting to protect the user at the MCS Colloquium

Facebook. Almost everyone uses the social network that changed the way people connect with friends. Unfortunately, posts on the social network may not be as secure as previously thought. According to Dr. Aaron Beach of the University of Colorado, Facebook can release enough information for strangers to figure out private information.

Beach received his doctorate in the field of Computer Science from the University of Colorado and now works on how to anonymize data that is put online. His lecture, titled “How Facebook is Anonymized,” attempted to solve the question of executing data management with concern to privacy, because how does one define personal information? “Is the size of your photo anonymous information? Is that personal information?” asked Beach.

Thus far, many cases have been made about what can be released. Google is facing charges for releasing a name of a blogger who said remarks that were considered offensive. Now Facebook allows for third party companies to look at personal information and then use the information for advertisements. “[The company’s] goal obviously is to provide personalized services,” stated Beach.

Facebook claims that they receive approximately 30 billion pieces of content (videos, pictures, status updates, etc.) a month. But since they do not post everything as ‘content,’ Beach estimates that Facebook receives more than thirty times their prediction per month. Beach said, “Who knows how many [that] would be, but it is probably something like a trillion objects.”

With the information that Facebook collects and releases, Beach was able to find that he could be traced to a single event. As he went through and saw that he was tagged by a friend who was taking pictures at a event, and the others who were also there all have a similar link to which ties all of them to a certain event in the past, which was a wedding in Panama for this example.

But Beach emphasized that the playful example could be twisted and used against any person or group. With enough information, one could discover personal identification information which can lead to identity fraud.

Beach talked about how there have been many attempts to release personal information without saying to whom the data belongs. But with a system of elimination, it is possible to determine a person with a unique combination of traits.

The example that Beach used was elimination by education, so of the people that went to his high school, undergraduate and doctoral institutions, and who were in Denver, he was able to find himself. Beach explained, “You have to do this weird thing, where you assume all of this data may be public and I have to treat it as if it’s private.”

The reason why the data is released is for statistical databases which need information to know what to produce. For instance, an increase in diabetics might require more research in that field. Beach’s forte is on web services, which release data for commercial use, such as advertisements that appeal to a person’s likes, and in cases of Facebook, the Facebook applications.

Thus far, Beach’s solution to the problematic personal information release in Facebook is to have a anonymized layer between Facebook and the applications, which filters what the application may find out about the user. According to Beach, the one wanting data cannot assume what the user would not mind allowing to be shown. Beach has developed a program that uses algorithms to find what two objects of information can be used so no one person may be unique.

Beach intends to use his application and hopes that it will go universal as it can help keep personal information private while still allowing companies and statistical databases access. For with Beach’s work, those who are using personal data for malicious intent will be foiled to find that they cannot pinpoint any unique identity.