Suffix Processing

kzinda · Post by **kzinda** » Wed Jan 15, 2003 3:30 pm

I am searching for concept variants of "sterilization" in my text. While it would appear to me from the morphene stripping routine described in the manaual that after stripping, the word form being matched in the text is "steriliz," Texis does not find the form "sterilized." Conversely, "sterilized" does not find "sterilization." Why is that? And yes, suffix processing is turned on.

Post by **John** » Wed Jan 15, 2003 3:47 pm

What do you have minwordlen and defsuffrm set to? They will also effect how the word is stripped.

kzinda · Post by **kzinda** » Wed Jan 15, 2003 3:52 pm

minwordlen is set at default, which I beleive is 5.

Post by **mark** » Wed Jan 15, 2003 4:08 pm

With default suffixes sterilized strips to steril and sterilization strips to steriliz. You could add ization to the suffix list to correspond to to ized.

kzinda · Post by **kzinda** » Thu Jan 16, 2003 11:45 am

I still don't understand why if "sterilzation" is stripped to "Steriliz", why it doesn't recognize "sterilized". Doesn't Texis look for the pattern match (steriliz) and then check for appropriate suffixes?

Post by **mark** » Thu Jan 16, 2003 11:56 am

When it finds sterilized it strips it down, then compares to what it's looking for. Basically, it compares roots after all suffix removal.

kzinda · Post by **kzinda** » Thu Jan 16, 2003 12:05 pm

SO what your saying is that when it finds sterilzed in the text, it strips the the word (steril)and compares it against the stripped word in the query (steriliz)?

Post by **mark** » Thu Jan 16, 2003 12:13 pm