I went to the University of Arizona, and as an undergraduate, I would sit in on masters- and postdoc- level classes where folks were discussing stuff like agrep.

Somewhere in my files I have the follow-up to this paper, which allows for affine weighting for various symbols. For example, you might say that vowels are more interchangeable than consonants, if you're looking for fuzzy matches in the pronunciation problem space. I think the author of that paper went on into bioinformatics in a big way after that.