Search This Blog

Tuesday, February 19, 2013

Facebook Graph and semantic search - what is that ?

Most of us are computer nerds addicted to Internet…. Many would search the next street in ‘google’…..   Not all realize that what you get from ‘google search’ need not be the best result or the right one at that…. Google Search (or Google Web Search) is a web search engine owned by Google Inc. Google Search is the most-used search engine on the World Wide Web, receiving several hundred million queries each day through its various services. The order of search results on Google's search-results pages is based, in part, on a priority rank called a "PageRank”

And most of us have a Facebook account and spend lot of time in posting, updating, and seeing FB.  Recently, on January 15, 2013, Facebook announced their new product Graph Search, which provides users with a “precise answer” rather than a link to an answer by leveraging the data already present on its site.  And those results are from content already posted by all FB users in public domain.  FB, the attractive social networking service was launched in Feb 2004 and now has over 1 billion active users, most accessing it on their mobile devices. For creating a FB account, one must have an e-mail account, through which they log in and then create a profile, add other users as friends, and exchange messages, including automatic notifications when they update their profile.  How many Friends you have in Facebook ?

The biggest drawback on ‘google search’ is that it provides results not for your intentions but for the key words that you typed on ~ Facebook Graph search claims that it  by semantic search, it  seeks to improve search accuracy by understanding searcher intent and the contextual meaning of terms as they appear in the searchable dataspace, whether on the Web or within a closed system, to generate more relevant results.  This search engine, as they claim, is designed to give answers to user natural language queries rather than a list of links. The Graph Search feature combines the big data acquired from its over one billion users and external data into a search engine providing user-specific search results. In a presentation headed by Facebook CEO Mark Zuckerberg, it was announced that the Graph Search algorithm finds information from within a user's network of friends. Additional results will be provided by Microsoft's Bing search engine.

This FB Graph feature was developed under former Google employees Lars Rasmussen and Tom Stocky and presently its beta version only is launched as a limited preview.  Graph Search operates by use of a search algorithm similar to traditional search engines such as Google, however the search would be based on intended meaning. Rather than returning results based on matching keywords, the search engine is designed to match phrases, as well as objects on the site. Search results are based on both the content of the user and their friends’ profiles and the relationships between the user and their friends.  It is stated that Facebook supports searches for  Public posts; people; pages; events; applications; groups; specified locations and the like. 

Perhaps you can make searches like : ‘those who like Sourav Ganguly” [or ……..] ; photo of my friends taken at Temples [or malls / cinema theatres].  The beta version for sure has generated lots of interest.  But first thing, Graph Search is only available via Facebook – there is no standalone app here.  So sitting at Office you cannot go to toolbar of the navigator and start searching in Graph search but has to log on to FB a/c first.   We are so habituated to open when you want to search anything [be it lost spectacles !] ~ of course the utility and probability of finding what we want drive the habit more.  For now, Google is unmatched. 

In Graph Search. Entries into the search box are autocompleted,  but that is nothing great for the regular user who is by now accustomed to this facility.  The search results are reportedly  easy to navigate and straight forward. The most relevant results will bubble to the top of the page, while the lesser relevant results, of course, get buried.  It is credited to have good filters which can whittle down the nearest of what you are searching for instead of providing lakh and more entries as answers to your search. 
books liked by my friends

At this stage, it is only a beta version and I was able to search : Photos of my friends and me; Photos taken in Chennai, of my friends; books liked by my friends.  Though FB pretends to offer greater variety, at this stage, not able to check whether it is really semantic and provides intelligent answers.
photos taken in Chennai of my friends

photos of my friends and me

There are lurking concern of private information unintentionally exposed here. some Techies quipped that  “people are now sharing horrible things about themselves thanks to Facebook Search.” It often seems as though Facebook’s main purpose is to remind us continually of how much we have chosen to share with the world about our online behavior—whether we realize it or not.  Whatever it be, it should provide a great deviation from the mundane : Status / Likes / interests / photos / friends / following more to more robust and interesting pieces of information overflow. 

Facebook already has good tools for analyzing content to get closer to consumers, and Graph Search would sure hold them more addicted at least for a brief while after full version introduction.. are you bitten by Graph addiction already ?

With regards – S. Sampathkumar.


  1. Homes for sale are comparable to other areas in the country in terms of numbers although the rate of
    sell far exceeds those in other places. That's not just farmers either, there are many in the city that utilize these for a variety of reasons. Depending on your specific requirements, you can have single, double or triple section steel barns or have one built to your particular needs.

    my homepage ... home depot sheds for sale

  2. Interesting.. I do have a FB account and would try FB graph today itself - Bavani