The PhD student didn't figure out the problem in the code over the weekend and we still weren't able to figure it out today. She decided that it isn't that important to the functionality of the program so she said we should move onto testing. She went and talked to my mentor and my mentor told her to test by keeping tally of how many false positives (predicted policy shows something not in the person's real policy) and false negativeness (predicted policy doesn't show something that is in the person's real policy) there are, and to keep track of the features that have these errors occur on them. This testing is to see if the program is accurate or not in predicting a privacy policy for someone, based on the group of people they are matched with that have similar features.

I wasn't feeling well today so I didn't work long. We started working on the policy comparison of users. I also started writing up the summary of the findings on the social networking site research.

We finished up the policy comparison part of the program and added more of the features to the code. We also tested the code on 128 users and noted the results. I finished up writing the summary paper and did research on one last site my mentor was interested in; Reddit.

Today I added one more group of classes to the code. It started processing very slowly at that point in time. We wrote down the results for 10 users before we had added that last class. My mentor told the student to create an automated function that would go through all the user tests over night because of how slow it is.

I met with my mentor and she liked the summary I had put together. We made some key observations about the social networking websites. She then asked me to summarize the similarities between 2 or 3 of the papers I had read that are most closely related to this work. I also need to go back and look at the types of joint-owned content is posted on the websites. I went to meet with the PhD student to work on the code - she had asked me to write in the connectivity to the rest of the 48 classes. She never showed up at the lab today. I finished doing all my work and went home for the day.

It also turns out all the work I did on Friday is to be scrapped. The PhD student had talked to my mentor about the slow processing time and my mentor said to only test for 3 of the 8 cases then so that's what we worked on.