I was doing some research on ANNs for categorization and the IBk algorithm came up. It was described as a k-nearest neighbor, but besides that, I can't seem to find anything about it. Does anybody know a paper describing this algorithm and how it differs from a regular k-nearest algorithm? Thanks!

The IB stands for Instance-Based. I'm not sure what it means in the context of kNN, since kNN classifiers are instance-based by default.