An Indian American artist-researcher has designed a data project that seeks to get a more accurate read on how voters feel by harnessing data from social media to overcome the shortcomings of the traditional polling methods.
The display at SPACES gallery in Cleveland, Ohio, gives an idea of how perceptions of each presidential candidate change over the course of the primary season — just like the content, positions, and tones presented by the campaigns themselves.
Roopa Vasudevan whose project, #Bellwether, focused on Ohio, examines social media data, and asks whether one can extract a “more nuanced or accurate read on the political desires of the voting population of Ohio,” a state that is regarded as vital to the endgame of both parties, but whose citizens, she says, “are often treated as faceless, nameless votes.”
Her project debuted ahead of the Republican National Convention, held in the city July 18-21.
Traditional polling tends to condense “the thoughts, opinions and preferences of voters into very specific, carefully curated statements” with small sample sizes dependent on the antiquated technology of landlines, Vasudevan says. Conversely, on the Internet, it is “possible to be overwhelmed by millions of different individual voices on social media, and they are often saying the same or similar things.”
Between August 6, 2015, when the primary campaigns began heating up, and July 12, she wrote a computer program to collect every tweet possible that mentioned any of the 22 presidential candidates.
After filtering them down to those relating to Ohio, she said, “I ran a variety of algorithmic language analyses on the content of the postings to determine a number of things: sentiment, frequency of similar or identical posts and phrases, grammatical structure, and the like.”
She shows her findings in a monthly timeline created from “remixed campaign material” that a visitor can visually understand. “These are actual, physical objects that I am creating with the social media data as the base. There isn’t a single screen present in the exhibition,” she said. “The design language — color, typography, imagery and logos — mirrors that of the official campaigns as closely as possible.”
Vasudevan is an Assistant Arts Professor of Interactive Media Arts at New York University in Shanghai specializing in using technology to explore and expose patterns in culture and behavior, and how computers and algorithms can help better understand relationships we have with ourselves and each other.
Last year she examined the environment in which split second reactions of momentous consequences happen in police-public interactions through an interactive digital project, “Hands Up,” that she co-created at the Flux Factory in New York City.
Vasudevan holds a BA in Film Studies from Columbia University, and an MPS from NYU’s Interactive Telecommunications Program. She is a member of the Flux Factory artist collective, based in Queens, N.Y.
The post Designing A More Accurate Measure Of Electoral Politics appeared first on News India Times.