Uncharted: Big Data and an Emerging Science of Human History by Erez Aiden
My rating: 4 of 5 stars
This nonfiction account of the creation of Google’s Ngram Viewer is fascinating. An Ngram is a word or phrase (N words long) and the Viewer measures how often that Ngram appears in books in recorded history up to 2008, at least in those scanned by Google. The authors devised the program’s basic features to view history and social change through a factual scientific lens, to see how our word usage changes over time and what that tells us. It begins with the example of illustrating when the United States changed from a plural to a singular noun. Popular accounts attributing that to the Civil War fully uniting the states into one entity once and for all turn out to be false. The trend toward the singular began before that and didn’t really take off until after 1880. If you don’t want to read the rather dry prose and the authors’ own speculations on social trends you can go directly to the appendix to see some of the charts that tell us how Santa compares to Satan, when data became more important (in books anyway) than God, and so on. They do touch on other forms of big data, but I wish they had spent more time and space on things other than Ngrams. What are the possible benefits and harm of all those photos being massively uploaded onto the Web? What about medical data – can it be used to identify causes or cures of diseases by examining massive trends. Google is now already quicker and better at predicting flu outbreaks than the NIH based on web searches for terms like”flu,” “influenza,” “fever,” etc.