Facebook has revealed details about Unicorn technology, the software and search engine which makes its Graph Search possible.
The technology helps the social network assign numbers to users, places and other reference points. Unicorn is an inverted index that can theoretically sort out queries with several operands.
Facebook search quality and ranking lead Sriram Sankar told The Register that without Unicorn, Facebook’s search technology would be very slow. "The Unicorn infrastructure strength is searching for entities based on attributes about these entities you can get. It is very different from a SQL database," Sankar said.
"Unicorn can be used to look up any table, like a hashtable can – but that’s not a perfect use of Unicorn. You’d use Unicorn when you have a bunch of keys."
The technology can index nodes such as anything with a Facebook ID, including a place, person, photo, post and ‘edges’, as well as allow linking objects such as check-ins, friendships, and tags.
Facebook will feed the Unicorn with huge data once in a week and the index consumes smaller bits of data all through the week.
When a user puts query on Graph Search, the list of search results is narrowed down by static ranking as per the importance.
Facebook’s new Graph Search will enable users to perform specific searches about friends on the site through key phrases.
Currently, the social networking company has no plans to make the Unicorn technology open source.
