Using social networks to gauge the mood of a community

world-wellbeing-projectOver the last few years there have been an increasing number of attempts to use social network data to understand various things about the real world.

For instance, only recently I wrote about a new project that is using Twitter data to try and underpin better urban policy making and planning.

Researchers at the University of Pennsylvania believe they can understand the well-being of a community via the Twitter posts of members of that community.  What’s more, they believe this data can help to guide a whole range of disciplines, ranging from medicine to national security.

“As people are more and more migrating toward social media for their social lives, social media increasingly becomes the platform for researchers to understand social trends, to understand psychological trends and to understand public health threats?Twitter ,” the researchers say.

The researchers believe that the words we use online provide a good insight into our personalities.  They can then use algorithms to go through our words and try and discern particular patterns.

“Social media increasingly becomes the platform for researchers to understand social trends, to understand psychological trends and to understand public health threats,” they say.

The study saw data harvested (with permission) from around 100,000 Facebook users together with around one billion tweets.

They began to notice particular words were common based upon the personality of each user.  For instance, extroverts would commonly use ‘party’, whereas introverts would refer to ‘computers’ or ‘Internet’ in theirs.

A better form of GDP

The researchers believe that such data could, and indeed should, be used by governments as a more cost-effective means of measuring the wellbeing of a society.

“The GDP doesn’t care about the nature of transactions in society,” the researchers say. “It’s one index among many. We’ve been working — and other people have been working — to convince governments to take on these other indices.”

Despite the apparent potential of the project, the researchers don’t believe that agencies will be switching to social monitoring any time soon, especially when the agency is heavily data driven.  They suggest however that security agencies may be more attracted to social media monitoring.

“This is small potatoes for the intelligence community,” they say. “The budgets that the intelligence community has thrown at population surveillance through these methods is to the order of a hundred, if not a thousand-fold, of what we will spend on this project.”

A major challenge for such projects of course is the reliability of the data.  A paper published last year in Science highlighted the challenges of getting unbiased data from social networks.

“Not everything that can be labeled as ‘Big Data’ is automatically great,” the authors note. There is almost a perception that if you have a large enough data set that any vagaries will even themselves out.

“But the old adage of behavioral research still applies: Know Your Data,” they say.

It’s particularly challenging when you consider the number of fake accounts that are littering the social landscape.  Whilst we are getting better at judging the real from the fake, it’s still likely that any data set will contain a large number of fake accounts.

Hopefully these kind of issues are being taken into account by the Wharton researchers when judging the well-being of their communities.

Related

Facebooktwitterredditpinterestlinkedinmail

Leave a Reply

Your email address will not be published. Required fields are marked *

Captcha loading...