|
Metamorph's whole search process is built around its ability to deal
with root morphemes in words, and in fact the program was originally
christened "Metamorph" based on that ability. An inherent part of
the search process is to take search words, strip them of suffixes and
prefixes down to a recognizable morpheme, search for the morpheme,
find a possible match, then go through the process again with the
possible match to see if it is indeed what you were looking for.
The elements which affect morpheme processing are very subtle as they
affect all language and therefore all of any text you are searching.
Therefore, any changes entered into the defaults which dictate how or
whether morpheme processing is done, will affect broadly the nature of
your search results.
The options which affect morpheme processing are the following:
- Content of the Equiv-Suffixes List;
- Content of the Prefix List;
- Content of the Suffix List;
- Prefix processing on/off toggle;
- Suffix processing on/off toggle;
- Morpheme rebuild on/off toggle;
- Minimum word length setting.
The content of the Equiv-Suffix List determines which suffixes
will be stripped before doing lookup for a matching entry in the
Equivalence File.
The content of the Prefix and Suffix Lists determines
precisely what will be stripped from a search word to obtain the
morpheme which is passed to the Search Engine. Specifically this
relates to all words in all equivalence sets.
To understand why the program is doing what it is doing you need to
know more than just what is contained in the lists; you must also
understand the sequence of the Morpheme Stripping process. If you
understand these rules, you will have better judgment on how to edit
the prefix and suffix lists, what happens when you turn off prefix
and/or suffix processing altogether, what happens if you turn off the
morpheme rebuild step, and where the minimum word length setting comes
in.
Copyright © Thunderstone Software Last updated: Sun Mar 17 21:14:49 EDT 2013
|