Analyzing Big Data Comes With Big Challenges

Dec 9, 2014

Credit Martin Gysler / Flickr

All those clicks, likes, and retweets you do on social media add up to a huge amount of data that can be worth millions to some companies.

But according to new research, actually using that data is easier said than done.

Derek Ruths, assistant professor of Computer Science at McGill University in Montreal, Quebec stresses that, at a general level, using social media data to predict behavior is challenging.

That's because data analysts are typically asked to study data for purposes different than what the data was originally intended for-- in effect, “trying to extract information that wasn’t intentionally put in there.”

Using data from social media platforms means doing things like negotiating with different user demographics and constructing populations based on imperfect or incomplete information. Because there's much that's unknown and unknowable, analysts of social media data should approach the data with caution, Ruths advises.