Graph Search will appear as a bigger search bar at the top of each page. When a user starts typing to search for something, Graph Search will make suggestions of searches users area looking to perform. Once a user selects the search it will return results on a page that a user can further customize and filter (for things like specific timeframes).
Facebook programmers have been working on the new search feature for about a year. The first version of Graph Search focuses on four main areas: people, photos (by location or topic), places and interests. The new search feature will facilitate searches of LinkedIn for jobs, Yelp for restaurants, Match for dates. Facebook has been collecting data on users, and is betting that friends will prefer recommendations from friends on restaurants, services and products, more than advice from a media food critic or a stranger on a message board. Initially the data mining that will be used come from users' pictures, likes, and check-ins, but not their status updates. Instagram photos will not initially be part of the Graph Search database.
For data outside of Facebook, Microsoft's Bing search engine will be resourced to help find information on the Internet.
Graph Search also potentially undermines the privacy protections in Facebook’s advertising model by indirectly reaching target population segments, which the American Civil Liberties Union believes can allow advertisers to easily identify Facebook users whose publicly-available data fits a specific profile.