Viktor Mayer-Schönberger and Kenneth Cukier, Big Data, A Revolution That Will Transform How We Live, Work and Think, London 2013
A short review for the Digital Humanites
The already vast amount of data doubles every two years. In Big Data Viktor Mayer-Schönberger and Kenneth Cukier describe the opportunities that lie in this abundance of hidden digital data that, so far, have been analyzed only to a small extent. Big Data is one of the most important current digital trends.
The book contains many interesting case studies. One of them deals with the prediction of the spread of the H1N1 infection due to the evaluation of queries in Google. Another surprising example is that of the UPS parcel service. In the USA, their drivers turn whenever possible to the right to get to their locations. So, UPS saves each year ten million dollars. They discovered this with a software that correlated accident statistics, information on fuel consumption and the records of the tours. The analysis revealed that UPS vans are much less likely to be involved in accidents if they do not turn left and thus cross the oncoming traffic, but instead turn right three times and then cross the road straight ahead.
Big Data refers to the use of algorithms for interpretation of large amounts of data, is the application of mathematics to large amounts of data for the calculation of probabilities. It’s about correlations and not about causality. The What is important in Big Data, not the How. To the essence of Big Data belongs the fact, that digital data do not have a pre-determined value, but that the combination of different data sets can create a new, unexpected surplus (value).
The focus is on data, „the description of something, that allows it to be recorded, analyzed, and reorganized”. To enable this, information must be dataficated, transformed into a quantifiable format. Such forecasts, which are derived from the past, must also interest the historical sciences. In a Virtual History (Niall Ferguson) the probability of historical events could be calculated, historical hypotheses tested empirically.
An example for the opportunities that Big Data arise in the Digital Humanities, is Google Books. Scanning books with an OCR results in dataficated texts, which can be searched with keywords. One application of this is the Ngram Viewer, that searches a huge amount of books. I see then, for example, when the use of the term “structure” outdistanced the term “event” (in the 1950s namely).
In this connection, the authors expect the construction of a new academic discipline: Culturomics – a computerized linguistics, which seeks to understand human behavior and cultural trends based on the quantitative analysis of texts.
Big Data is an important book for the Digital Humanities. Reading the chapters, Correlation and Datafication is highly recommended for anyone who deals with data-driven research. In other chapters, the authors also focuse on the risks of a data driven society. This will not be discussed here.
Note: Please find here the link to an interesting post of Frédérique Clavert on the Digital Humanities and the Datafication: Mise en données du monde, mise en données de l’histoire?