The UK Terror Watchdog says this is very difficult in practise, is it?

Question

I saw this quote today in the Guardian

David Anderson QC, the terror legislation watchdog commissioned to report on the state of Britain’s surveillance laws in the aftermath of Snowden’s disclosures, has previously said that internet connection records included storing details of every website visited up to the first forward slash in their address – but not a detailed record of all web pages on a particular site. The internet connection records will also include times of contacts and the addresses of the other computers or services with which the user made contact.

“Under this definition a web log would reveal that a user had visited eg google.com or bbc.co.uk but not the specific page,” said Anderson in his report, A Question of Trust. “It could also of course reveal ... that a user has visited a pornography site or a site for sufferers of a particular medical condition, though the Home Office tells me it is in practice very difficult to piece together a browsing history.”

Given this information logged over time (it's kept for 1 year) by an ISP, would it really be difficult to piece together the browsing history of an individual? (ignoring the other types of information that is either already available to them or they can get further to this)

The only hurdle I can see is differentiating the individual from others at that physical address, but if an analysis of Facebook likes can reveal if I'm gay then I reckon a year's worth of browsing history from one physical address could easily be divvied up into individuals. Even if the number of individuals is unknown.

What have I missed?

---

Edit: I wrote to David Anderson QC, and he tells me that is a quote from his report and is only stating what the Home Office told him, his opinions are given later in the report. I don't think this changes the validity of the question though, but I don't want to be unfair to him.

Answer 1

Would it really be difficult to piece together the browsing history of an individual?

I think this is considered metadata. In my opinion, metadata can piece together what a user was doing:

They know you rang a phone sex service at 2:24am and spoke for 18 minutes. But they don't know what you talked about.

They know you called the suicide prevention hotline from the Golden Gate Bridge. But the topic of the call remains a secret.

They know you spoke with an HIV testing service, you doctor, then your health insurance company in the same hour. But they don't know what was discussed.

Perhaps not totally on topic, but I hope you get the idea.

Answer 2

If you want to reconstruct fully detailed browsing history (eg the exact URLs of pictures the invigilated user fapped to) then the answer is: no, it's not possible at all from this logs alone.

Such logs give information on where to look for such detailed data: They know the source (invigilated user) and target (the visited site) address, so They can try to subpoena the target site into providing full logs identifying exact pages visited by given source address. Such narrow request (only about one user) is often more successful than broad request (give all your logs about everything). It contains only small piece of the information so the owner of the site often perceives as not threatening. (Eg if he was to give full logs, those logs could be used to tell which pictures are the most popular - something very useful for competing sites). So yes, such logs make it much easier to get fully detailed history.

However, as Jeroen said, your question is moot by definition. The metadata is the most important data itself. Metadata is far more useful for invigilation and they are several orders of magnitude cheaper to analyze than direct eavesdropping. Metadata pretty much is already compiled statistical picture of target behaviors and preferences, ready to be applied, while real data (eg. phone tap records) requires time consuming work of trusted and expensive humans to understand and compile into a bigger picture.