Page:Note-by-Chair-and-selected-documents-ordered-from-Six4Three.pdf/249

 I was most interested in the sessions on big data applications, and privacy and security (http://bgdata csail mit edu/annual2013). The speakers in the applications session spoke about applying data to public health problems, identifying genetic diseases, and improving education using MOOC data. It's great to hear positive applications of data that motivate solving the privacy questions. Each of the speakers noted that their research has been impeded by unsolved privacy questions like how do you obtain informed consent, and how do you design a system with transparency in mind. In the privacy session, Danny spoke about designing for trust when users might fear surveillance. He proposed that accountability must be embedded in the system design so that questions about data provenance, use, and transfer can easily be answered. Here's a neat demo on visualizing public Tweets. http://mapd.csail.mit.edu/

Maritza also attended the first meeting of the Big Data privacy working group. The group plans to approach the problem using specific case studies. The tentative plan is to focus on the MOOC data set first, and walk through the process of applying the available technical solutions and policy guidelines to define an arrangement that would allow the data to be shared with researchers. As a subtask of this work, the group will decide on a set of assertions that describe a successful data sharing program.

Maritza met with Alan Davidson to discuss the proposed MIT policy initiative.

Maritza, Rob, and Erin continued to iterate on a proposal for a research program centered on the privacy paradox Maritza continued to work with Cam and Bob Kraut on a proposal for a data donation program with Wolfram Alpha.

-- Maritza

From: Erin Egan Date: Friday, November 15, 2013 2:02 PM To: Matt Perault .Nicky Jackson Colaco . Maritza Johnson Cc: Rob Sherman Subject: FW: HPM Thanks Rob. Others? From: Rob Sherman Date: Friday, November 15, 2013 2:00 PM To: Erin Egan Subject: HPM Log In Anonymously. Rob continued work with the Open Graph team on Log In Anonymously (name subject to change), a product that will let people use their Facebook accounts (or, potentially, another Facebook-mediated account) to log into apps without the need to provide personal information or consent to sharing of data App post suggestions. To address concerns about apps posting "implicit" user actions back to Facebook timelines, the Open Graph team is working on a feature that will, on an opt-in basis, enable apps to send "suggested posts" describing your activity to a "locker" within Facebook. If enabled, users will be able to see potential actions they could post to their timeline and then tap on a suggestion to open up a pre-filled composer. Rob is working with the team on technical implementation details to address privacy concerns HIGHLY CONFIDENTIAL FB-00089883