> MZ> For my mind, i strictly recomemded to make russian affixes for both
> MZ> cases, because one part of webpages may contain 'yo', while other - 'e'.
>
> And how could an indexer determine what of they to use for each page?
By normalyzing. It search all possible prefixes and suffixes. For
'yo'-pages it will find 'yo'-affixes, for 'e'-pagwes - will find 'e'
affixes.
> And how could search engine determine have it to use
> 'io'-containing-dictionaries for user request analizing or not?
It should contain both. This not for spellchecking, this for
normalyzing.
And 'yo'-'e' equivalence of normal forms may established by synonym
ruleses (this in TODO).
--
Maxime Zakharov http://sochi.net.ru/~maxime/
Sochi, Russia http://www.sochi.com/
___________________________________________
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]
