Agencies have a wealth of unstructured data — images, audio recordings and other information that doesn’t fit neatly in traditional databases or lend itself to analysis by traditional data tools — at their fingertips. So how can government make sense of all this data? How can agencies actually use it?