I'd like to add a stemmer token filter to the subjects field and apply non-quoted queries against that. Variations on the same word would be caught in non-quoted queries. In quoted queries against the field, the full word will still be required.
This is to reduce the need for * wildcards. Someone who's searching for Architecture might also be interested in items with architect subject tags, but the field isn't great at capturing those at the moment. The analyzed field would allow us to accommodate a more specific tag-like structure and the broader utility of a full-text search field. Prefix match ranking would still place more relevant items above.
I'd like to add a stemmer token filter to the subjects field and apply non-quoted queries against that. Variations on the same word would be caught in non-quoted queries. In quoted queries against the field, the full word will still be required.
This is to reduce the need for * wildcards. Someone who's searching for Architecture might also be interested in items with architect subject tags, but the field isn't great at capturing those at the moment. The analyzed field would allow us to accommodate a more specific tag-like structure and the broader utility of a full-text search field. Prefix match ranking would still place more relevant items above.