Hello Dear People,
I have a few questions about the use of GiST for text search.
My situation is this:
I have a large and rapidly growing database of documents, and I want to support text search.
The documents contain lots of words that it would be valid to search over. I think this means a GIN index will become too large, and a lossy index is
I've tried setting up a GiST search (just using the default English configuration). The performance is disappointing.
The GiST/text-search documentation says "GiST indexes are lossy because each document is represented in the index by a fixed-length signature."
Is there a way to configure the signature which GiST + tsvector is using?
Presumably a longer signature would produce a larger but faster index?
Where should I look to find more information on what exactly GiST is doing for the text search case?
Thank you in advance for all help!