algorithms to generate word forms

John Milton (lcjohn@uxmail.ust.hk)
Fri, 21 Feb 1997 15:58:51 +0800 (HKT)

Can anyone tell me whether there is a publicly available set of algorithms
that
1) stems, i.e., produces real-word lemmas from the inflected and derived
forms and
2) generates the derivational & inflected forms from the base form.

Ideally (but not necessarily) it would report word class (e.g., differ=v;
difference=n), when an orthographic form is instantiated in more than one
word class (e.g., supply=n/v) etc.

I know that this problem has been investigated for quite a long time, and
that there remain some fuzzy edges to it, such as polysemy and, as Adam
Kilgariff, for one, points out, there is no widely accepted rule for what
forms fall within a word family (e.g., does differ include difference &
differential?).

Has at least a workable solution been found that does not generate a
lot of errors and is it accessible?

Thanks, John
__________________________________________________
John Milton
lcjohn@uxmail.ust.hk
The Hong Kong University of Science and Technology