Per Thorsheim tweeted the following:
Some people took notice, and he thought he should explain the idea a bit better, in case somebody would actually try to take the challenge. As always, possible fame, mentions & maybe a cup of coffee awaits those who can put something together.
1. Background
Ashley Madison was hacked. Data was dumped. And a user known as @deuszu seemed to know more about it than most on Twitter. Douchebag, Smartass or the real deal? That is best left for others to properly investigate. In any case, Thorsheim was fascinated when he saw that @deuszu mostly wrote new tweets, even when commenting or replying to others. Without watching carefully and using the search function, the tweets look like an endless rant from someone who should drink less coffee. Brian Krebs also made a note of this in his post on August 15: Who Hacked Ashley Madison?
2. The Challenge
There are tons of Twitter analytics tools out there. Forensic toolkits as well, such as TwitterForensics Maltego had something for Twitter as well. But given the rather unique nature of deuszu's tweets, it wasn't clear how much they could help. Here's the most basic form of what was needed:
- Enter a link to one tweet from a specific user, or enter a username
- Show replies to that user or particular tweet (or show the threaded conversation if there is one, including if the user replied to someone)
- Allow basic search filters to see replies to, or mentions received by A) the user's followers, B) accounts the user follows and/or C) anyone else
Now for the parts that probably make this new & unique:
- A date/time threshold option to see only E) mentions of this particular user from [selectable from A,B,C) and/or F) tweets (tweet/mention/reply) made by [selectable from A,B,C)
The purpose of this is to find tweets from the target user that are indeed comments/replies to others who may have mentioned them, or that they follow. Bonus option: searching for keywords used by them (input list) that are also used by others.
Example:the user tweets "Tango down!", and one could search for "Thunderstruck|ACDC|tango|down|dump" in tweets from that user's followers, or anyone they follow.
3. Why?
It's unclear whether such a tool would ultimately be useful, but it certainly could be. Someone could patent it, set up a company called "tweet conversation forensics Ltd", sell the product for a ton of cash, get rich, stay faithful to their spouse and go on vacation to Bermuda. Thorsheim noted he wouldn't care much, as long as he received a free license covering his needs. He would prefer it as open-source though, being an idealist who wants to help make the world a better place.