• View
• Edit
• History
• Print

Privacy

1. Suppose I have one data set that lists the names of student at UW, the year they entered UW, and their hometown. I have another data set listing towns and the standardized test scores of students at UW from those towns and the year the exam was taken. Explain the privacy risk: who is at risk of having their test scores divulged and why?
2. You want to address this using differential privacy. Instead of publishing either data set, you instead allow users to submit queries asking about students, hometowns, and test scores. How would you apply differential privacy so as to not leak any private information?