Tracking down Twitter retweets... a fuzzy approach

Question

i know that tracking retweets in this fashion:

Source -> User1 -> User2

is not possible as Twitter only gives you the original tweet and not any retweet of a retweet. Based on this.. do you think it might be possible to track down retweets based on followers of the originator of the tweet? For example, what I want to do is to get a retweet, find all users that have retweeted it and then find how those users are related to the person that had originally tweeted it. If B is not a follower of the originator but have retweeted a tweet… could it be that a person B that he follows is a follower of the originator of the tweet and B retweeted it so A eventually saw it?

Thanks in advance for any feedback.

Answer 1

I am currently looking at how a user interacts with their timeline, and this is a problem I have considered as well. The problem is that retweets are not always a result of a tweet appearing on a user's timeline. The user could frequently retweet tweets found in the public stream, trending topics, or the "explore" tab.

I think it is possible that your approach could work, but you would need to apply some filters to the kinds of users you look at. You would have to limit the amount of followers you look at, as many retweets come from very popular users. Crawling 10,000 users simply to track one retweet isn't really an option considering the limits of the API.

A more viable option may be to track a retweet from some user that retweeted it. For example, say you have User A, who retweeted a tweet that originated from User B. You can look at all of the users that User A follows, find any candidates that have retweeted User B's tweet (including User B itself), and then search the users they follow, etc, until you reach User B. It would resemble a depth-first-search (or breadth-first, depending on the implementation details). However, you could run into a problem if any retweeting user along the search is private. In this case, you will not see their information and the path of retweets is broken.

I know this isn't a black-and-white answer, but hopefully this helps a little. It's an interesting question.