News360 says it does recommendations using semantic filtering, etc. Washington Post launched a service called Trove that it says will do the same — and aspects of recommendations appear in other apps as well, and the New York Times recently launched its own recommendation service attached to its site.
These kinds of recommendations are largely algorithmic, although the NYT has some behavioral info that comes from anyone who has joined its internal Times social network, which was an interesting experiment — although it’s not clear how many people are actually using it.
The biggest source of recommendation-type data is Facebook, which can see when your friends like something, when they share it, etc. Huffington Post has driven a lot of traffic to the site by making smart use of Facebook integration to recommend stories that people you follow have liked or read, and as I keep pointing out to people — social networks like Twitter and Facebook are inherently farther ahead when it comes to recommendations because of all the social signals that are embedded in my social graph, the relationships with the people I follow and my friends and social network.
There are some services and apps that make use of the links that get passed around via Twitter and Facebook — there’s Twitter Tim.es and Paper.li for showing you links from your Twitter stream, and PostPost does something similar for Facebook links, producing a kind of personalized newspaper. News.me, the social news platform that Betaworks and the New York Times have partnered on, shows you content from other peoples’ streams, which is an interesting twist. And Flipboard pulls in links from your Twitter stream, your Facebook graph and RSS feeds and shows it to you.
But no one is really doing much when it comes to recommendations. I’ve tried playing with Trove, and it is not much better than a random sampling of news that is being shared on the web — and News360 seems equally haphazard. It’s possible that they could get better over time, of course, although there doesn’t seem to be any way to tell either service that it is wrong when it suggests a particular story to you. And News360’s choice of a visual interface with photos sliding by is interesting, but I’m not convinced it’s particularly useful.
Others are trying to solve the recommendation conundrum, including **, which former ** is developing as a kind of **. But so far, if you want recommendations about what to read, your Twitter stream and Facebook graph are probably the best solution — and anyone who wants to do better is going to have to leverage both of them to do it.