- Category: February 2015 - Data Driven Marketing
Despite all of the promising indications of a rosy, data-driven future, there are also critical voices that urge not to overvalue Big Data and to ensure correct handling of it.
These voices include, for example, Professor Richard Rogers, professor of New Media and Digital Culture at the University of Amsterdam. He presented five points of criticism regarding Big Data at the Data Days in Berlin, basically concluding that we are at present in a phase of "disappointment" after the initial hype, but with Big Data getting a clear upward trend again.
First point of criticism is the potentially hasty "reputational shift to large-scale infrastructures and their analysts", a change of image and budgets towards large infrastructures and the few people that are able to handle it.
The second change Rogers observed, he called “paradigm shift away from the interpretation (hermeneutics) of text and data towards pure pattern recognition, from close reading to distant reading", meaning from the detailed consideration of the individual aspects towards a holistic view.
This would lead to the third phenomenon, namely the possible apophenia, the false detection of patterns where actually there are none. Since the highly complex graphics and analysis of Big Data could only be read by a few experts, only this specific group would be able to make statements, which could lead to over-interpretations.
The fourth point in Rogers lecture takes up the same issue again, criticizing the fact that the decision makers have been reduced to small specialist teams, saying that we are "relying on one small team of data scientists with special data access privileges". Common analysts and experts would rarely be consulted.
As a final point of criticism, Rogers argues that Big Data may mislead a false sense of comfort. There is a "shift to relying on available signals rather than seeking out unavailable ones", he said, meaning that you should rather rely more on elaborate analysis of existing data, then to try to search out new data that might not even be available yet, but could potentially be more relevant. As a consequence people would just search and not tackle the actual problems lying ahead.