verification (expert system)

Terms from Artificial Intelligence: humans at the heart of algorithms

Page numbers are for draft copy at present; they will be replaced with correct numbers when final book is formatted. Chapter numbers are correct and will not change now.

Verfication is essential if an expert system is to be accurate and trustworthy. The internal consistency of rules (whether they agree with one another) may be assessed algorithmically, but the external consistency (whether the rules make the right choices in the real world) is harder requiring expert assistance critiquing the ruels as they are gathered or additional expert labelling if potentially problematic/inconsistent cases are found.

Used in Chap. 18: page 297