While all IT leaders who work with data to improve analytics understand that accuracy is important, guaranteeing that accuracy is not always a simple pursuit. The problem is that no one has a uniform, blanket rule that can be applied to all companies and their information management strategies.
Everyone has a different goal when delving into "big data." For some companies, the aim is to dominate a global economy and make billions, while for others, there are smaller aspirations, such as getting a better handle on a small neighborhood market. Both goals are equally valid, but it's worth noting that these two different companies will have two different ways of looking at data quality.
As Information Management recently put it, it's a matter of "rotating frame of reference." Cathy O'Neil, data science consultant at Johnson Research Labs, likened data quality to astronomy. If you're watching the Earth's rotation from here on our home planet, you have a different perspective than anyone who's anywhere else in space.
"It depends on your frame of reference," O'Neil said. "If I'm standing on the Earth, and I look up in the sky, I will observe the Sun going around the Earth in a wobbly path. Although it would be quite a bit simpler to understand the model of the solar system whereby the Earth and other planets revolve around the Sun and spin while they do so."
In astronomy, it's vital that we understand context. The sun may look wobbly, but that's really just our skewed perspective. In business intelligence, the same principle holds true. Data may paint a certain picture for an enterprise, but that picture comes into better focus when data scientists interpret it better and place it into clearer business context.
In other words, companies should think about their particular goals. What are they really looking to accomplish through analytics. And how, specifically, can a given cluster of data make that happen?
