Best Practices - Fluid Topics - 3.6 - Technical Notes

Understand Linguistics in Fluid Topics

Fluid Topics
Technical Notes

Stemming dictionaries must only include:
  • Singular or plural forms, like green = greens
  • Masculine or feminine forms, like in French, vert = verte
Each equivalence within these dictionaries must only cover one token.

Example of an invalid dictionary

Stemming: (1 to many | many to many)
  • horse = smart and beautiful quadruped
  • car wheel = car wheels

Example of a valid dictionary

Stemming: (1 to 1)
  • horse = horses
  • car = cars
  • wheel = wheels
Other vocabularies must only include:
  • Synonyms, like green = emerald
  • Variants, like neighbor = neighbour
  • And/or generic and specific terms, like means of transportation > car = automobile > SUV
Note: Synonyms, taxonomies, and thesauri should never include inflections: inflections must be added to stemming dictionaries only. When adding a new entry in the thesaurus, you must verify that inflected forms are correctly defined in stemming dictionaries to fully benefit from thesaurus expansions.