The Case for Auto-Classification: The Content, Context & Attributes

A critical aspect of successfully managing electronic records is to be able to take massive amounts of unstructured content and separate items to buckets for action.  Manual classification of electronic records is even more unreasonable than it was with paper because, with computer-assisted classification, items often can be put into more than one class – like a duck billed platypus, for example, which could be a mammal and a bird.

Some digital objects are what they are because of what they discuss; others because of what function they serve.  To successfully identify and classify electronic records into exclusive categories, on a massive scale, one needs to leverage the content, the context, and the attributes together, and be able to do so in a repeatable and defensible way.

If you are struggling with this lets chat. Modern tools and techniques allow for metadata classification, subject based analytics, predictive coding, near duplicate analysis, and topic modeling to achieve effective classification.