//CGCS Media Wire is proud to present a inside look at Truthy. Karissa McKelvey, and Filippo Menczer of the Indiana University Center for Complex Networks and Systems Research discuss the motivations, inner workings and practical application of the API power house aimed at helping researchers, citizens, and journalists understand how information spreads on social networks.
BY: Karissa McKelvey and Filippo Menczer
Information travels at unprecedented speed with the maturation of the Internet as the primary cultural and communications medium. Throughout the history of media technology — the printing press, newspapers, radio, movies, and television — we have seen the speed of communication increase steadily with each advancement in technological capability for long-distance communication.
But the Internet is different.
The Internet, and social media in particular, allow citizen journalists to record events in real-time and spread information faster than we’ve seen with traditional media. More importantly, social media are catalysts for
facilitating the creation of communities within vast communication networks, connecting millions of people around the world. But how do we cope with this flood? How do we separate genuine facts from spam, misinformation, and astroturfing? Who are the most influential actors, and how do they acquire such power? How do people use this new media to organize and form opinions? How does the Internet affect politics, economics, culture, and society? Thankfully, social media don’t only enable us to connect; they also enable us to explore these questions.
With each keystroke, we leave behind digital trace data — each of us has a digital footprint. Social media such as Twitter have open-access APIs that make it possible for researchers to collect data about users and their connections in aggregate. With access to this data, social scientists can study human social behavior on a larger scale than has ever been possible in human history.
However, there is a catch. Collecting, managing, and analyzing massive amounts of data from social media is a challenge, even for expert computer scientists. On the other hand, the qualitative analysis of rhetoric, argument, knowledge, community formation and organization theory cannot be simulated in a computer program (yet). Collaboration between computer and social scientists is key.
Bridging this gap is one of the research goals of our group in the Center for Complex Networks and Systems Research, at the Indiana University School of Informatics and Computing. Truthy is a project aimed at helping researchers, citizens, and journalists understand how information spreads on Twitter. We currently focus on tweets about politics, social movements, and news from the past 90 days. The site monitors political activity to build interactive visualizations that allow visitors to identify the most influential users, popular topics, and interesting events. Users can see information about the online activity of information spreaders, such as their sentiment, language, and political leaning.
We recently unveiled a timely interactive visualization tool focused on Twitter activity related specifically to the upcoming US presidential election (truthy.indiana.edu/elections). Designed to benefit not only researchers, but also journalists and voters, the election coverage tool charts tweet sentiment and volume related to presidential candidates and political topics, annotating the highs and lows in the timelines with content from popular links shared on Twitter.
We also visualize the diffusion networks of election memes, showing how these memes spread from person to person. Users can then explore the features of the key spreaders. With these analytics, one can begin to understand how information spreads through the social network, asking question such as: How does sentiment change over time in response to political events? What is most popular over time? Who are the most influential users on a particular topic?
Other tools on the Truthy site let users view timelines of volume for individual memes, map the geographic diffusion of these memes, generate YouTube movies that display how hashtags emerge and connect, and download related content directly from Twitter. Currently we are working on the technical challenges of new features such as automatic truthiness detection, crowdsourced fact-checking, and an open-source API. We’d love to hear about research you’d like to do with our data, so that we can design the API with you in mind. Please contact us with feedback, comments, requests, and bugs.
We hope that our interactive visualizations and open tools will help bridge the gap between social scientists, data miners, and the public, to jointly advance our understanding of online social networks and their role in information diffusion, especially in domains such as politics, where social media have enormous power to affect the electoral process and consequently the policies that impact our lives.
//Karissa McKelvey is an Informatics PhD student at Indiana University. Research interests include Complex Systems and Society, Information Diffusion in Complex Networks, Information Visualization, Social Media and Political Process, and Computational Social Science. Currently, Karissa is a graduate research assistant at Indiana University’s School of Informatics and Computing under the supervision of Filippo Menczer and Fabio Rojas.
Center for Complex networks and systems research:
Indiana University informatics and computing: