Unstructured data introduction
Unstructured data vs structured data?
What’s the difference between structured and unstructured data? If we look at structured data – that which you’re most familiar with, it’ll be similar to that of an excel spreadsheet. You’ll have columns and rows. Most of us are very used to that and we can navigate it relatively easily. The metadata in the column headings also helps us understand the content that sits within the rows.
Unstructured data are things that are more like images, audio or text where there’s no structure there that allows us to interpret the data easily and we have to use different techniques in order to unlock value from it.
The health care industry is a great example of why we need to do more with unstructured data. It’s estimated that around 80% of the medical record data that is captured about us is unstructured – things like handwritten physician’s notes, scans and the various different medical images that we might be getting in X-rays. Google has given a great example of how you can use this kind of data to drive better outcomes for, in this case, patients. They took some mammograms and other scans that looked at that medical images and using AI with it, they were able to detect breast cancers more quickly, more accurately, and with fewer false positives than trained medical doctors can do.