IBM: Dark Data and How Cognitive Computing Figures It Out
This podcast was originally published in Cloud Technology Partners
We discuss dark data and how IBM Watson can understand it for decision-making. Normally when people refer to data they think of structured data in rows of numbers. But in the last few years, the data has changed into images, videos, medical scans, sensor scans, audio and telematics. This dark data is unstructured, which makes it difficult to analyze with existing systems and therefore harder to use for decision making.
Currently 80% of the world’s data collection is dark data, and that’s expected to grow to 90% by 2020. So if you’re running a data insights-based company, you need to start analyzing this unstructured data now. IBM Watson can read the unstructured data, and it make meaning out of it. Imagine a home security camera that recognizes the UPS uniform and lets you know when your package arrives. This is an example of machines assisting humankind, not competing with it.
Facebook, Uber, and Google are fast-moving companies that are becoming the new definition of “Enterprises”. Traditional enterprises are trying to keep up with their speed. Looking at the innovative businesses, one common thread is that they all have a huge investment in cloud and in AI. It’s not just about machine learning and deep learning. Ultimately if you can have machines that understand the user in their native tongue, when they need it most, and understand the urgency, the needs and the emotions at a given time, it will help you make critical decisions based on human context. That is a powerful differentiation for any enterprise in the marketplace.