Facebook’s Graph Search: The Semantic Web’s First Steps into the Mainstream?
by Nick Iannitti, Fuel Industries / February 8, 2013
Note: This blog entry is available in English only.
Last month, Facebook Announced a new feature, Graph Search, that allows users to search through their connections in a way that, well, actually makes sense. Type ‘coworkers who like sushi’, and you’ll know who to ask out for an after-work dragon roll. Type ‘photos before 1990′ and you’ll easily find that picture of you and your buds waiting in line for Whitesnake.
On one hand, it’s a way to keep people on Facebook for longer, and, just maybe, run a couple fewer searches on rival Google’s engine. Hey, if I can find a decent Mexican restaurant recommended by my friends, without needing to leave my Facebook chat, then why wouldn’t I?
On the other hand, it brings to mind an interesting question in regards to the ‘non-quantifiable’ data in our lives: In, say, a decade, what will have more ‘truth’: the truth that is relevant to one’s social sphere, or the webpage with the highest ranking on Google? This is the challenge of the semantic web, the idea that the internet can actually achieve an understanding of not simply which documents or pages we are searching for, but what we actually want.
Tools like Google Voice and Siri attempt to remove certain intermediate steps from a user’s query, trying their best to fill in the blanks and ‘decide’ how to best help, while sites like amazon or last.fm try to cater a user’s experience to their apparent tastes, but it’s still a shallow system. And while we’ve yet to see exactly how deep Social Graph’s capabilities will go, social networking is such a major slice of why people are online in the first place, that comprehending how humans interact with other humans is a key ingredient to computers being able to assess our true goals and eventually grow to serve us more often, more quickly, and more fluidly. (If you’ve got 3 minutes to spare, this is a great little primer of the concept.)
So we go social – we hook the internet into our friend list, into our photo feed, into our likes and dislikes, and see what the computer spits out. For now, as the staff at Mashable shows us, the insights that Graph Search can provide aren’t exactly earth-shattering (did you know KFC is the most popular food for fans of Star Wars?). As soon as Facebook’s tool goes fully live, with the new ability to search through their profiles and past updates, many users will immediately be floored by what is representing them online. But this will change. Both Facebook and others will continue to refine the tools, and as users begin to have more sophisticated interactions online, and the ‘internet of things‘ takes hold, the social space will be more reflective of real tastes and real insights into a user’s thought process.
An example: a 10-year-old female Taylor Swift fan and a 49-year-old male fan of Phil Collins both run a search for ‘best albums of 2012′. Their world views and life experience are completely different, as is their understanding of what the ‘best’ music might be. Should their search result be identical? On one hand, one might argue ‘yes, how else will people discover new things and broaden their horizons?’ One the other hand, is there really any value to telling the 49 year-old that Rihanna’s latest album made the top 50? And vice versa for the 10 year old and Rick Ross’s new EP? In reality, both of these people were actually trying to make a playlist for a roadtrip – but the search engine couldn’t know that because they didn’t choose to type it. But with access to what they’re saying to friends, imagine: instead of Google’s ‘did you mean’ function correcting spelling, it actually corrected meaning? ie. Did you mean you’re looking for great roadtrip music for your drive this weekend?
We’re pretty sure that tackling a question like ‘what is truth’ is a bit too ambitious for a blog post, but it’ll be interesting to see the internet put in an earnest effort in the years to come. Ethical and privacy questions are fodder for a whole other discussion, of course, but in the meantime, I won’t be surprised if, in a few years, we’ll be looking back and wondering how we ever managed having to click through search results to find the best taco in town.