Hi Tom,

I've created a new Oracle Text index with a user defined datastore. The index works, but I'm having trouble understanding how Oracle Text handles plural vs. singular when searching. For example:

SQL> insert into foo (col1) values ('elephants');

SQL> select count(*) from foo
where contains(col1, 'elephants', 1) > 0;

COUNT(*)
--------
1

SQL> select count(*) from foo
where contains(col1, 'elephant', 1) > 0;

COUNT(*)
--------
0

Is using the 'fuzzy' operator the best (or only) way to address this:

SQL> select count(*) from foo
where contains(col1, 'fuzzy(elephant, 70, 6, weight)', 1) > 0

COUNT(*)
--------
1

or is there something else that can be done when creating the index? I've tried using a wordlist preference with the STEM_FUZZY_PREF attribute set, which I thought would obviate the need to use the 'fuzzy' operator in the contains clause:

begin
ctx_ddl.drop_preference('STEM_FUZZY_PREF');
ctx_ddl.create_preference('STEM_FUZZY_PREF', 'BASIC_WORDLIST');
ctx_ddl.set_attribute('STEM_FUZZY_PREF','FUZZY_MATCH','ENGLISH');
ctx_ddl.set_attribute('STEM_FUZZY_PREF','FUZZY_SCORE','60');
ctx_ddl.set_attribute('STEM_FUZZY_PREF','FUZZY_NUMRESULTS','5000');
ctx_ddl.set_attribute('STEM_FUZZY_PREF','SUBSTRING_INDEX','TRUE');
ctx_ddl.set_attribute('STEM_FUZZY_PREF','STEMMER','ENGLISH');
end;

but that doesn't have any effect. BTW, what does the 'contains' clause in the search functionality of your asktom application look like?

Thanks much!