Abstract
Large web-based data sets make possible political studies at a scale inconceivable just a decade ago. Everything from personal opinions to popular political movements leaves a footprint online and provides a firsthand account of both everyday and historic events. These data are, however, unstructured and full of noise requiring new approaches—quantitative methods developed in the realms of political and social science, but also in data analysis and mining. Applied to online data, these make possible language modeling, topic tracking, novelty detection, social network mining, and many more types of analyses, all providing new insights into social and political realities.
The works compiled below are a sample of approaches to quantify political phenomena using online data. All articles underwent a peer-review process by experts from both computer science and/or political/social science. The topics covered include general-purpose social media like Twitter (Crooks et al., Dyagilev et al.) and blog posts (Sagi & Dehghani), news outlets (Yom-Tov et al., Bright & Nicholls, Sagi & Dehghani), search query logs (Yom-Tov et al.), governmental websites (regulations.gov in Levy & Franklin), and political citizen-driven sites (debatepolitics.com in Liang). The findings in these works illustrate the richness of the data available online, and the tools one may employ to quantitatively assess political opinions expressed there, the use and readership of online content, and the various social networks people form, as witnessed through their actions online.
We believe that these scientific articles present timely research in an area that will continue to grow and draw more attention. Hopefully, this special issue can contribute to make this research area more interdisciplinary and initiate new discussions and collaborations.