Real-time search engine “OneRiot” is now indexing Facebook Likes as part of the Open Graph API:
[W]e were very happy to see Facebook opening up public data with the Open Graph API and other products at f8 last week. For example, “Likes” can now be shared across the web in interesting ways. Realtime search engines can now process that “Like” as a social signal to help increase the relevance score of a page in the index.
At OneRiot, we have now started to experiment with this Facebook data.
Until today, we’ve been indexing the links shared on Twitter, MySpace, Digg, Delicious and by our own OneRiot panel to help determine our search results. Now, with the addition of Facebook data, OneRiot delivers search results that reflect the pulse of a much, much wider social web.
That’s very nice — using Likes as a relevance signal for the algorithm — but there’s something more interesting here.
I had previously thought that only Facebook would be able to access all the Like data and wrote about what that might enable in the local segment:
All the local publishers in North America and Europe implement Social Plug-ins, etc. Millions upon millions of businesses are “Liked.” That creates a master data set in the aggregate. Facebook will know:
- The favorite sushi restaurants in every city
- The favorite contractors
- The favorite attorneys
- The favorite . . . in every category
They’ll know all these things in the aggregate and in terms of my network in particular. Each “Like” is a “vote” in the same way that each link is a “vote” in the SEO world. But a Like vote is much better than a link vote.
In a very short period of time Facebook will have a ton of valuable data. What will do with all this data (courtesy of all the local sites that will implement Social Plug-ins)?
I’m not a developer or engineer so I’m on shaky ground a little bit here. But if, as OneRiot says, all the Like data is available to third parties then they can do what I was suggesting only Facebook would be able to do: create a killer database of local favorites.
You can bet that Google will be thinking about this and tapping into that data if it’s public. Shouldn’t you also?