Integrations Journal



Facebook Graph Search Marries Big Data With NLP

Written by  Harpreet | 05 February 2013
E-mail PDF

facebookFacebook recently launched its upcoming feature, currently in beta, called Graph Search. The new feature allows anyone to perform a search on Facebook using simple text based queries. These differ from a conventional web search you normally do on Google or Bing in a way that you're not really entering keywords here.

If you were to search on Facebook's Graph Search, you'd normally type - "My friends who like beer", "My friends who like beer and Apple", and so on. Facebook's Graph Search is a unique blend of big data and natural language processing (NLP).

We had earlier reported how Facebook's Graph Search is a lean way to put its enormous amount of data to good use. With a billion users under its belly, Facebook is one of the most interesting use cases of how big data evolves with a social graph. Facebook is slowly taking control of much of our lives. We are no longer dependent on Facebook to communicate or just connect with people, we're now looking at it to help us find relevant things we want. In the process, we also expect it to return suggestions along the way.

Using natural language processing (NLP), Facebook is slowly trying to bring users to a different side of the search world. Till now we were all used to enter random keywords, or at the most - writing complete sentences - on Google, but not anymore. Facebook, if successful, is set to change the course of history by being the more "human" search engine. It doesn't really compete with Google's web search as of now but sometime in the future it will. NLP will help it distinguish itself from conventional search engines in a number of ways, while also providing comfort to its users.

Data is the center of every search function. Facebook's chewing up on a massive scale, even by big data standards. It's one of the biggest places to handle consumer data, almost all of it highly private and confidential. Facebook's deep concern right now is to keep that data secure and away from eyeballs that don't deserve it. It's probably one of the reason why the new Graph Search functionality is being made available to a select few users right now. Facebook works on a highly successful iteration development. They keep building new features, testing them on select users and then rolling it out eventually to everyone after taking the feedback into consideration.

Search is expected to grow in the coming years. Information, especially social, is mostly with a select few popular social networks and Facebook is the biggest of them all. It'll be interesting to see how market dynamics shift in the search industry, in the coming few months. Will Facebook spend as much as it needs on infrastructure this year? Will they come up with more big data and NLP implementations on the world's most popular social network? We can't put a confirmation tag on that yet but the signs obviously suggest a positive response.

[Image Credits: Facebook]



Harpreet is a technology journalist based in India. He currently writes on Mobile, Technology and Startups. He is an avid reader and a passionate writer. Prior to ToolsJournal, Harpreet used to write for a major English news daily.

blog comments powered by Disqus