RE: [Snowball-discuss] Stop word lists

From: Martin Porter (
Date: Tue Oct 08 2002 - 20:40:01 BST

Alex (and Oleg),

I was asking about the origin of the lists to see whether we could advertise
the Finnish list without infringing copyright.

Certainly stopword lists are database specific, in the sense that optimum
performance of a database IR system will be achieved by a stopword list that
is dependent on the characteristics of the IR system, but we should also
remember that it is possible, for any language, to build up a list of
'neutral' words, and that that list can be used in a general sense to help
IR performance.

See the notes on stopwords in the introductory account of Snowball.

Anyway, the lists themselves are proving very useful.


This archive was generated by hypermail 2.1.3 : Thu Sep 20 2007 - 12:02:43 BST