Skip to Content

Review: Big Data: A Revolution That Will Transform How We Live, Work and Think

Printer-friendly versionPrinter-friendly version
Fall 2013

John Porter (VCR)

“Big Data” poses major challenges in perspective for information managers and scientists alike. The book “Big Data: A revolution that will transform how we live, work and think” by Viktor Mayer-Shönberger and Kenneth Cukier does a good job of laying out opportunities and challenges afforded by “Big Data.” There are some aspects of their advocated approach, specifically the emphasis on correlation over causation, that represent major challenges to our science. To quote the book: “The ideal of causal mechanisms is a self-congratulatory illusion; big data overturns this.”  Nonetheless, the power of “Big Data” is, and will, influence how we manage and analyze data within ecology.

The book is written for the non-technical reader and is replete with a wide array of real-world examples of the application of “Big Data.” The chapter headings of the book “Now,” “More,” “Messy,” “Correlation,” “Datafication,” “Value,” “Implications,” “Risks,” “Control,” and “Next” say a lot about Big Data by themselves, emphasizing the velocity and magnitude of data collection, how when you have lots of data, you can live with less accurate or precise data, the power of prediction, and the opportunities and challenges of Big Data.

As elsewhere, the definition of “Big Data” is a sliding scale, being defined more by the need to apply non-traditional analytical approaches than the size of the data itself.  A strength of the book is the wide variety of examples used. Many of them come from the world of massive data derived from Internet search engines, or web crawlers (e.g., Google’s ability to predict flu outbreaks based on search terms used), and social media. However, there are also examples that are less anthropocentric, such as the application of many low-cost, low-accuracy sensors in place of a few high-cost, high-accuracy sensors.

A thesis of the book is that many of the rules that applied in the past no longer apply today. For example, when you can collect essentially all the data from a population, sampling becomes unnecessary. Similarly, when you have really massive amounts of data, irregularities in merely large amounts of data become inconsequential.

As a scientific reader worshipping at the altar of accuracy and precision, there was much in the book that seemed radical or even subversive.  However, although I don’t think that “Big Data” approaches will entirely supplant traditional sampling approaches, they can augment and enhance our ability to address ecological problems, allowing us to attack problems that are not soluble using traditional scientific approaches.