[comp.text] Word similarity criterion?

conrad@cgl.ucsf.edu (Conrad Huang%CGL) (04/14/88)

A friend of mine is writing some software for handling bibliographic
references.  One of the things he wants to do is to compare two references
and see if they are "the same".  The problem is that the references may
actually be the same but one may have a typo in it.  What he really needs
is some criterion for determining whether two words are similar.  Could
anyone point him to some papers, or better yet, some code?  Thanks,

Conrad

UH2@PSUVM.BITNET (Lee Sailer) (04/16/88)

See the SOUNDEX algorithm.  Knuth, The Art of Comp. Programming, Vol. 2
is a good reference.