I have done three similar projects, the first is for creating a social network in which I created a social network of a user and find the loopholes in the user account with respect to Network Analysis. The second project I created is for Twitter, in which I also extracted tweets of a user. I also used a scheme to download more than allowed user profiles. After downloading I predicted the behavior of a user. The thirds project was done using SELENIUM in which I had 1000 email accounts of different users, I had to create a BOT that reads all the accounts and filter emails according to their content. The system use to read 5 emails at once. The system was able to run on GMAIL, HOTMAIL / Outlook and it use to avoid security.
I have an idea to make your project avoid in a legal way, this is what I did in my Twitter Project because I had to download millions of tweet and user profile but the limit was around 300-500. I downloaded more than that figure but within the limit. I have shared you some link in which you can see my Twitter dataset (this is not the full dataset its just a sample) and SNA network analysis.
[login to view URL]
[login to view URL]
[login to view URL]