Rank features allow us to store sparse vectors and weights within the index. Rank features store the 'term' being indexed and the corresponding weight, which represents the importance of the term.
Rank features are used for semantic searches underpinned by SPLADE type models.
Within the Control Hub go to Mappings > Rank Features.
Select Create New Ranks Feature.
Enter a unique ID for this Rank Feature.
This must be in camel case and only contain alphanumeric characters and underscores.
Enter a user friendly name within Name.
This must be alphanumeric characters and underscores only.
Explain what the rank feature is for within Description.
Select Create Rank Feature to finalise.
