Hi,
Did you check on Lucene's website that Lucene does not have what you are looking for? Which would be less of case here.
Stemming makes use of Porter's stemmer algorithm to figure out stems of words. I am sure you are already aware of what stemming is. Considering you need to implement your own stemmer for the targeted language extending logic of Porter stemmer's algorithm for English.
You can find the base alogorithm here
This implementation is in java so you may have to some extra effort to conver to C# and then integrate this analyzer into kentico and build your index using this analyzer during indexing process.
Does this hlep you? Let me know.
Thanks,
Chetan