Subscribe

Categories

Add to Technorati Favorites

 Subscribe in a reader

Trends and Outliers

TIBCO Spotfire's Business Intelligence Blog

02/14
2011

Data Visualization Weeds Out Social Pollution on Twitter

truthy1 300x211 Data Visualization Weeds Out Social Pollution on TwitterIn our #followfriday series, we look at how social BI and data analytics professionals used Twitter during the previous month. Twitter has become a source to get a pulse on the latest ideas and topics of most interest to the community at large. We create hashtags to represent ideas and tweet and retweet these ideas to raise awareness and create a discussion.  Unfortunately, some of the buzz on Twitter isn’t real – it’s fabricated. But a group at Indiana University is using data visualization of social network connections to identify and weed out the fabricated “social pollution.”

Stephen Colbert coined a phrase called a “truthy” – misinformation that is promoted as fact using deceptive tactics.  A truthy could be a hashtag, mention, url or phrase that is passed among several Twitter users, many of whom are most likely fictitious people created  specifically for the purpose of propogating the truthy. The Indiana University Center for Complex Networks and Systems Research has created Truthy, a project dedicated to identifying and revealing these fabricated discussions. The group uses text analytics, sentiment analysis and social network analysis to create visualizations that make it easy to spot fabrications.

This network diffusion visualization illustrates what is almost certainly a truthy. The two black dots represent two different Twitter users who do not disclose their identities. The top dot has made more than 10,000 tweets over the past four months, all in support of several political candidates. The tweets include links to websites that support the same candidates. The second black dot has not made any original tweets, but instead has retweeted everything that the top dot tweeted, promoting the same candidates. By doing this, these two accounts created a “Twitter Bomb” –   a Google search of the candidate names will return these tweets, which are positive endorsements of the candidates, in the first page of search results.

truthy2 300x211 Data Visualization Weeds Out Social Pollution on Twitter

Indiana University Center for Complex Networks and Systems Research

This network visualization shows a clique of Twitter users promoting a video on the Internet. The thick blue lines highlight a group of very active accounts within the clique that each send thousands of tweets mentioning each other and promoting the video. In addition to tweeting about the other accounts and the video, each of these active accounts retweets every tweet from every other account in the clique, creating a giant, potentially fabricated buzz.

truthy3 300x211 Data Visualization Weeds Out Social Pollution on Twitter

Indiana University Center for Complex Networks and Systems Research

Twitter Spam is a series of accounts tweet about a certain website and include popular hashtags to draw more attention to the tweets. Each account tweets a different page in the website, and as you might expect, each account retweets everything that every other account tweets about the site. The accounts are quite possibly bots set up by the spammer to automate this social pollution.

Subscribe to our blog to stay informed on data visualization and other data analytics topics.

Steve McDonnell
Spotfire Blogging Team

Print post

Is your data hiding something?
Download Your Free 30-Day Evaluation of TIBCO Spotfire®

Post a Comment

Your email is never shared. Required fields are marked