Stemming dictionaries are used to store inflected forms of words present in a given corpus: it includes for each entry which is the regular form, a list of inflected forms, or variants.
In this case, you do not have to worry about gender (masculine or feminine) and number (singular or plural) anymore. This feature is integrated by default, but a few terms may be missing and can be added this way.
A document indexed using a stemming dictionary thus match any form declared in the stemming dictionary.
- The query "a brown dog" match documents containing "brown dogs" if "dog" = "dogs" in the stemming dictionary.
- The query "success stories" match documents containing "success story" if "story" = "stories" in the stemming dictionary.