Sunday, March 24, 2013
Tool Kit: Facebook’s Graph Search Makes Use of Friends and Likes
In January, Facebook began testing its new search tool, an enhanced version of the search box at the top of the site. The tool, which Facebook calls Graph Search, gets its name from “social graph,” a technical term for the giant network of connections among friends, friends of friends and so on. The social graph includes not just members’ names, but also the pages they have liked and the places where they have checked in. Graph Search lets a user concoct short phrases, instead of stand-alone search keywords, to search Facebook; for example, “books my friends like.” The tool does have limitations; if you’re looking for the nearest sushi restaurant in a big city, or trying to browse the complete works of the author Susan Orlean, searching Facebook is not the way to do it. In many cases, though, Graph Search lets you take advantage of the clicking and typing your Facebook friends, and their friends, have already done. You can start a Facebook search by going to facebook.com/about/graphsearch in a desktop or laptop browser, and clicking the big button at the very bottom of the page to replace the search box at the top of Facebook with the enhanced Graph Search box. Facebook has not made the tool available on mobile apps yet, but says it will do so later this year. Once the tool is installed, try typing a description of what you want into the search box, preferably using the term “friend” or “like.” Pages will pop up, as well as suggestions for searches based on what you have typed so far, using conversational language. This differs from the keyword-oriented style of Google, in which a user tries to think of words associated with what is being sought. Graph Search looks at Facebook’s organized data — names, likes, friends, locations, photo tags and so on — but not status updates or comments. FIND A BOOK One of the most obvious graph searches is “books my friends like.” That will display the Facebook pages created for specific books on which friends of yours have clicked the Like button. Moreover, Facebook will show you which specific friends of yours liked each book, so you know who suggested which one. You may have to scroll down a page or two to find a book you have not heard of before, or to dodge the collective bad taste of the social graph. (My own friends are too fond of books about social networking.) One way around this is to search for “books liked by people who like what I like.” This both broadens the search beyond your friends, and narrows it to those who have clicked Facebook’s Like button on the same pages as you. Some authors, like Ms. Orlean, are also active Facebook users who click the Like button on book pages. That means you can search for “books liked by Susan Orlean” to see more than a dozen reading suggestions from the author of “The Orchid Thief.” FIND A RESTAURANT Facebook cannot match Yelp’s extensive listings, reviews and block-by-block location targeting for finding a nearby restaurant. But while Yelp has a Facebook-like friend network, chances are you have not built up a network of your own there. Instead, you have built up a much larger network of Facebook friends, many of whom have been clicking the Like button on restaurant pages. With Facebook search, you can look for “restaurants my friends like in Boston” or “nearby sushi restaurants my friends like.” As with books, Facebook will tell you which friends like which restaurant, so you know whose advice you are getting. Keep in mind that Facebook’s idea of “nearby,” unlike that of a phone app, is not GPS-targeted. It can mean anywhere within city limits, even if it’s 45 minutes away. Facebook search works best for advance scouting of new restaurants worth traveling to, like “Japanese restaurants near New York, New York, liked by people who live in Japan,” or “restaurants in San Francisco, California, liked by people who like what I like.” In New York City, you can specify individual neighborhoods like “Park Slope, Brooklyn, New York.”
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment