11 Oct 2005
CS 5244 - Computational Document
Analysis
22
Signature methods
For text documents
¡Checksum
¡Keywords
¡N-gram
(usually character) inventory
¡Grammatical
phrases
For source code
¡Words, characters and lines
¡Halstead
profile
(Ignores comments)
lOperator
histogram
¡e.g., frequency of each type sorted
lOperand
histogram