Among the many millions of users on Twitter there are surely a few thousand who publish the content we are looking for. Getting smart recommendations is not easy, the typical algorithm of tell me who you follow and I will tell you what interests you It is not always effective, so it is important to keep looking for recipes that help create links in this content network.
Currently there are many systems that offer recommendations, but few of them make them in such a direct way as the experiment that Twitter discussed here. It’s about the account Magic Recs (@magicrecs), an account that sends direct (private) messages with recommendations from new users to each of their followers.
The system used to make the recommendations is not very original, although it does not have to be less effective. It is in charge of looking for common contacts among the people we follow to recommend those who are outside our circles.
For example, if a large percentage of those who read on Twitter read the account at the same time @wwwhatsnew, and you do not follow it, you will surely receive the recommendation of that channel.
As commented in The New York Times, the activity of this account may be expanded in the future, testing new recommendations, new channels of communication between Twitter and the followers of @magicrecs.