On Saturday, June 11, 2011 1:23:47 PM UTC+1, Adam Green wrote: > > I believe that at least part of Twitter's motivation is based on > protecting users from spam and viruses. In that case, why not > implement some purging alogrithms? Here's an easy one. If an account > follows nobody and only sends emails with a screen name and URL, it is > probably a spammer and the URLs are probably bad news for anyone who > clicks them.
Far from being easy - that's exactly the sort of pattern that twitter has come to realise they can't automatically classify as spam as it's exactly the model of usage to get news out in oppressive regimes (esp the "arab spring" of last 6-9 months). Imagine you've just seen the army roll in and start a massacre of civilians, you take a few photos, post them somewhere like a dropbox account with a short cry for help, set up a twitter account and rapidly post the url to a load of people outside the event - journalists, activists etc (note that you set up a new account so that it can't be traced to you should the army or secret police or whoever get hold of it). This model of rapid anonymous dissemination of breaking news is a major part of what makes twitter what it is - maybe not to all of us, but it's a key model that they want to foster. So what twitter do instead is wrap the link in a t.co link and expand it again when people view the tweet.. they can then examine the url after the event (ie a few seconds or a few minutes later) and if it turns out to be spam, disable the t.co link (so anyone clicking on it after that point will not be sent to the pointed at site). If you look it up you can find interviews with the twitter spam team where they explain this. -- T -- Twitter developer documentation and resources: https://dev.twitter.com/doc API updates via Twitter: https://twitter.com/twitterapi Issues/Enhancements Tracker: https://code.google.com/p/twitter-api/issues/list Change your membership to this group: https://groups.google.com/forum/#!forum/twitter-development-talk
