[Snowball-discuss] Multiple languages

Michael Wetherell mike.wetherell@ntlworld.com
Thu Feb 20 04:19:02 2003


On Wednesday 19 February 2003 5:46 pm, Vineet Gupta wrote:
> The best option would be to segment the text by
> language, and then run each through the appropriate
> stemmer.  It is fairly straightforward to do the
> segmentation, and running through multiple stemmers is
> likely to produce bad results.

I'll take your advice, I've pretty much figured out how I'll handle the details now. Multiple stemmers was a bad idea.

Thanks,

Mike