Even if data was continuously sent, you could still analyze the data by protocol. You can find out who sends e-mails to who, who our contacts are in Twitter, Fa

Even if data was continuously sent, you could still analyze the data by protocol. You can find out who sends e-mails to who, who our contacts are in Twitter, Facebook, Skype, how often we exchange data with them... One could build an anonymization tool that adds randomness addressing those one by one. For instance, to avoid our e-mail exchanges being analyzed, this tool could exchange random e-mails in the background with people we don't know. Just to mess up with who our contacts are. To make a list of "users of Tor" being no longer useful, the tool would send and receive data to the Tor network one random day per week. It would anonymize searches (this already exists) by making random searches. It should visit random websites But would you add random contacts to your Facebook account? To Skype? ... Dealing with all aspects is hard, it would waste electricity and resources, and I think it would just make analysis harder, not impossible.