Big Data (2013) by Viktor Schonberger and Kenneth Cukier is a somewhat breathless account of ‘Big Data’ which is the term for applying statistics and machine learning techniques on large datasets to find knowledge.
With the massive amounts of storage available, the internet and the proliferation of cheap sensors the amount of data humanity collects is growing rapidly. Coupled with gigahertz processors and parallel processing analysis of data sets of an unprecedented is now possible. Internet searches, phone numbers called, purchases made on credit cards, number plate locations and far more is now available to analyse.
Such data doesn’t need to be sampled and can be investigated in full. This provides considerable value to firms like Google, ebay, Farecast, Facebook, UPS and others to use their huge datasets in new ways. Google translate, for instance, has created an unprecedented translation facility using huge data sets where correlations are simply sought out rather than being added by deliberately translating word after word.
The book is not a bad read, but it’s not great. The examples in the book are interesting and it’s well written but the over excited tone detracts from an interesting tour of a new and useful technology.