Facebook Graph Search: making connections via things
Facebook Graph Search is rolling out on English language versions of the social network following a six-month beta phase. The social networking giant has built a network of more than 1.1 billion individuals (March 2013) connected via networks such as personal relationships, families, schools and organisations. It now wants us to explore the network and connect via things that we are interested in.
A simple search toolbar belies a powerful semantic search engine.
I can use the search bar to create queries around physical things such as books, films, music, places and restaurants. If Facebook can’t find an answer it serves content from the web instead.
Graph Search is intended to give you precise answers based on content shared by your network of friends, and friends of friends, using plain language queries.
For example I’ve just turned up photos of my village in Northumberland, a list of people that work in digital roles for a competitive agency, and musicians liked by people that have similar interests to mine. The results are remarkably good.
When you hover over the Facebook search bar Facebook serve suggested searches that you might want to try. I’d also recommend you try the ‘discover something new’ button that is served at the bottom of the right hand side bar after you’ve made an initial query. Both of these features enable you to explore Graph Search.
Inevitably there will be concerns about privacy.
It takes seconds to build lists of people based on their likes, location or relationship status. Facebook is seeking the counter this with user education such as online prompts advising users how to withhold their data by locking down privacy setting.
Semantic search is a vision of the internet of people and things as old as the web itself. Facebook’s execution on such a massive scale is a first. It’s a big data project in every sense of the term.