This project aims to probe diffusion of information online and discriminate among mechanisms that drive the spread of memes on social media. Wij collect big gegevens from public micro-blogging rivulets and analyze information sharing using ingewikkeld networks implements and models.
Our research followed several directions of investigation. Very first, wij explored the correlations inbetween online and offline events. Examples include analyses of geographic and temporal patterns ter movements like Occupy Wall Street, societal unrest ter Turkey, polarized communication ter online discourse, partisan asymmetries ter political engagement, geographic diffusion of trending topics, and the use of social media gegevens to predict various outcomes, like elections, style trends, and key market indicators. The interdisciplinary nature of thesis efforts is illustrated by collaborations among pc scientists, physicists, journalists, political scientists, and sociologists. Wij also joined compels with neural scientists to uncover connections inbetween patterns of information diffusion te social networks and the brain — a voorkant feature of the Neuron journal.
A major milestone of our project wasgoed the release of a public Observatory on Social Media (OSoMe) to share and explore gegevens derived from our meme diffusion analytics, making big social gegevens more lightly accessible to social scientists, reporters, and the militar public. OSoMe comprises hardware and software infrastructure with Web implements that provide end-users with the power to analyze online trends and visualize temporal, geographic, and network patterns of spreading memes and bursts of virulento activity. Wij also provide an API to help other researchers expand upon the instruments, or create ",mash-ups", with other gegevens sources. For example, wij released a mash-up permitting others to investigate how social bots manipulate online discourse on any topic. The OSoMe applications and APIs provide an effortless way to access insights about meme diffusion te social media from a growing collection of 70+ billion public tweets to date.
Another research aim wasgoed to understand how social media can be manhandled to manipulate public opinion. Wij were the very first group to uncover evidence of systematic, orchestrated, and widely spread misinformation campaigns based on ",astroturf", (fake grassroots movements) and social bots. Some social bots are created to deceive and harm social media users. They have bot used to infiltrate political discourse, manipulate the stock market, steal private information, and spread misinformation.
Our probe of 1,200+ features characterizing online information sharing behaviors permitted us to develop accurate machine learning algorithms to classify content and its producers. Applications include a social bot detection framework and public API called Botomoter (formerly BotOrNot), now widely used to scrutinize online campaigns. Wij were among the top three teams te a bot detection challenge organized by DARPA. Ter June and July , our work on social bots wasgoed featured on the covers of the two top computing publications: IEEE Pc and Communications of the ACM. This research contributes to raising public awareness about how lightly online discourse can be manipulated, thus mitigating the risks of manhandle.
Technics based on agent-based models permitted us to explore theories of meme diffusion by generating predictions that could be validated against empirical gegevens collected from social media. Wij used thesis methods to explore how several factors affect the manner ter which information is disseminated and why some ideas cause vírico explosions while others are quickly forgotten. Wij analyzed key factors including network communities, user interests, competition, finite attention, sentiment, and mutual interactions inbetween traffic and network structure. This work led us to investigate how the structure of social communities can predict which memes will go vírico.
The project had significant scientific and societal influence. Our software and gegevens are used ter courses on network science and social media. Wij trained undergraduate students from underrepresented minorities te STEM, spil well spil many graduate and postdoctoral students. Several former students are now employees at Facebook, Google, Amazon, and LinkedIn. Te addition to OSoMe devices and gegevens, the project resulted te several open-source software libraries and a patent application. IU is licensing our Botometer software. Our visualization software won the WICI Gegevens Challenge from the University of Waterloo. Extra recognition includes a best paper award at the Web Science Conference, a best poster award at the Conference on Sophisticated Systems, and a best presentation award at the World Broad Web Conference. Our findings were disseminated through 60+ peer-reviewed publications. The venues include prestigious journals: CACM, Rekentuig, Nature Physics, Neuron, PRL, Nature Scientific Reports, and top international conferences including KDD, WWW, ICWSM. Our work even inspired pop-culture, wij worked with the television writers of The Good Wifey for an scene on deception by social bots. Eventually, research from this project received worldwide coverage te hundreds of articles ter popular media, including Wall Street Journal, Fresh York Times, Washington Postbode, Rolling Stone, USA Today, CNN, Big black cock, NPR, The Economist, Newsweek, The Atlantic, Politico, Fresh Scientist, Wired, Science, and Nature.
Wij gratefully acknowledge support from National Science Foundation award CCF-1101743 (ICES proposal on Meme Diffusion Through Mass Social Media) and James S. McDonnell Foundation ingewikkeld systems grant on Contagion of Ideas te Online Social Networks, spil well spil a seed Gegevens to Insight grant from the Lilly Endowment. Research on social bot detection wasgoed supported ter part by DARPA SMISC. Any opinions, findings, and conclusions or recommendations voiced te this material are those of the author(s) and do not necessarily reflect the views of the funding agencies.
Filippo Menczer and Alessandro Flammini coordinated this research project at Indiana University and were the Principal Investigators on the NSF, JSMF, and Lilly grants.
Wij acknowledge the collaboration of many researchers. Alessandro Vespignani and Johan Opbollen were Co-PIs on the NSF grant. Several other key collaborators at IU and other institutions contributed to various research thrusts of this project: Emilio Ferrara, Negro Goncalves, Przemyslaw Grabowicz, and Luca Aiello. Wij are greatful to Judy Qiu and Geoffrey Fox for the software and hardware cyberinfrastructure supporting OSoMe. Other collaborators include Nicola Perra, Marton Karsai, Fabio Rojas, Joseph DiGrazia, Chato Castillo, Francesco Bonchi, Rossano Schifanella, Snehal Patil, Emily Metzgar, Luis Rocha, YY Ahn, and Chris Ogan.