identifies and extracts features/attributes from each digital artifact. The mechanism by which features are picked and interpreted depends on the approximate matching algorithm. The representation of this collection is the similarity digest of the object.
Source(s):
NIST SP 800-168