In order to create our recent election tweets map, we downloaded all geocoded tweets published from the United States in October: about thirty million messages!
In addition to extracting content containing election-related terms, I also wanted to see what the raw data would look like. i.e. where are these tens of millions of opinions, ideas, and statements coming from?
The map above, which simply consists of dots on a page, belies the intensity of effort that went into creating the image. My laptop spent days crunching, processing, visualising, and exporting.
You may look at the map and ask why. Is there anything really that surprising here? The volume of information simply seems to reflect human settlements and interactions.
But that is precisely what interests me about the map. Transactional data and short published snippets of information can never replace the richness and nuance that can be obtained from other modes of inquiry. But there remains something enthralling, captivating, and compelling about working with these data. Thirty million different human interactions recorded and visualised.