MaryAnne Gobble writes that the definition of big data is multifaceted, ambiguous, and dynamic, suggesting the term has different meanings between academic disciplines and practitioners. Put simply, big data sets are too big because of their volume, speed of information or velocity, or because of the variety in the type of data. Big data sets provide researchers with large amounts of information on many topics, ranging from demographics and psychographics to social networking statistics. These sets are often made usable and archived through data mining techniques.
Thomas H. Davenport, Paul Barth, and Randy Bean identify the many forms of big data, including “clickstream data from the Web, social media content (tweets, blogs, Facebook wall postings, etc.) and video data from retail and other settings.” Big data also refers to any digitally archived information, ranging from recorded content from telephone call centers to biological data from laboratories. Because the term big data refers to so many different types of data sets, standardization of meaning and analysis is rare across disciplines.
Big data collected from social media include the recording of user-generated content such as Facebook posting, tweets, user preferences, shared links, and profile information. Big data sets also typically record other less obvious measures, such as time spent lurking, deleted content, and geotracking locations. This information is stored in large servers and databases, and is sometimes only accessible with the owner of the social media's permission.
Barbara Trish declared 2012 the year of big data in politics based upon the way the U.S. presidential campaigns of Barack Obama and Mitt Romney used large data sets to connect with voters and influence public opinion. The Obama re-election campaign used big data sets derived from Twitter and Facebook to personalize messages, track issue efficacy, and target specific segments of the larger U.S. population. The Romney campaign used big data to identify potential donors and supporters based upon social media information such as purchasing history and religious affiliation.
Maurice Vergeer writes that new innovations in data collection and interpretation represent new possibilities in understanding the role of the Internet in politics. However, not all researcher perspectives support such an optimistic view. While the information sciences have made great advances in the interpretation and use of big data, traditional social-political science approaches still favor small-data and theory-driven approaches. Vergeer proposes the real advancement in social and political research will only occur when the two oppositional approaches work together and support each other's inquiry.
The development and integration of big data in the 2012 election also invoked debate between strategists about the potential future of politics. The collection and documentation of private information makes up the majority of concerns regarding the ethics of big data. If social media data is private, archiving and collecting this information is an invasion of one's privacy and can be deemed unethical. Alternatively, if social media data is public, archiving and collecting such information is ethical.
Alternatively, debate surrounds access and use of big data in politics. Big data sets derived from social media giants such as Facebook or Twitter are costly, meaning that political campaigns with smaller budgets are often unable to purchase the information. What the data sets are used for is also of concern. Joe Turow writes that information gathered from big data sets is beginning to be used by online retailers who adjust prices to meet specific socioeconomic information about individuals. This means that pricing is becoming more individualized and customized to what retailers believe an individual should be charged. Turow also notes that news Web sites are also starting to customize the display of news articles to match what they believe individual viewers will most likely read. Specifically, political campaigns use big data sets to create personalized or targeted advertisements to appeal to users. Just as news media that personalize headlines to reflect what they think will appeal to viewers, language, images, and messages are being customized based upon the information gained by big data sets. Turow suggests this is problematic because possible inconsistencies may develop from such customization.
Critics are also skeptical due to the size and volume of big data sets. Vergeer writes that there is a tradeoff for volume over accuracy that plagues the collection and use of big data. In order to analyze such large data sets, generalizations about social media indicators such as relationship status and political orientation can potentially be misinterpreted. For example, big data sets cannot differentiate when a child is browsing the Internet on his or her mother's computer. Therefore, there is an acceptance of big data analysts that there is inaccuracy within the information collected.
The future of big data collected from social media for politics is uncertain, although the success of its use in the 2012 U.S. presidential election suggests that it is an important tool in campaigning. As John Timpane reflects, social media and big data sets help politicians find votes and turn potential voters into supporters. Big data is a tool and likely a new frontier in politics and campaigning because of the vastness of the data it can provide.
See Also: Data Mining; Mobile Media User Data Collection and Privacy Protection; Obama for America iPhone Application; Open Data
Big data is the use of specialized computing techniques to locate patterns in and derive meaning from large and complex data sets. One of the most ci
See blogosphere; facebook; global scrutiny; mobilization; social media; twitter; youtube.