2. Information classification
As information will get housed in information lakes and different more and more related methods, one other problem is classification. Who’s allowed to take a look at specific information? From authorities safety classifications to confidential HR info, information shouldn’t be accessible to everybody. Information should be correctly categorized, and people classes and the boundaries they entail should be maintained and dwell on as corporations combine and harness information in new methods.
3. Stability
Quite a lot of information is transient. Should you’re taking information from sensors, for instance, it’s worthwhile to perceive how usually you’ll refresh the information based mostly on sensor readings. This is a matter of knowledge stability, as consistently altering information could result in totally different outcomes.
Information can also be growing old. For instance, think about you had a selected course of for elevating a job requisition for a brand new worker for 9 years, however you revised the method final yr. Should you use all 10 years’ value of knowledge to coach a mannequin after which ask the way to open a job requisition, more often than not, you’ll get a incorrect reply as a result of a lot of the information is outdated.
