[Snowball-discuss] Ukrainian stemmer

Martin Porter martin.porter@grapeshot.co.uk
Tue Nov 11 14:03:01 2003


Myroslav,

What you report is very interesting. 

I found that with two closely related languages it is convenient to use one
as a starting point for the other, even though the resulting stemmers may be
quite different. So to some extent the Portuguese stemmer developed from the
Spanish one, and the Norwegian from the Swedish. You might therefore find
the Russian stemmer a useful starting point. Use a test vocabulary about the
same size as the ones on the Snowball site. (It is a mistake to use a very
big test vocabulary - you'll find it just slows you down.) 

If you formalise the rules I could code it up in Snowball,

Martin