Reverse searching order; start with the presumably smaller specific set and...
Reverse searching order; start with the presumably smaller specific set and work our way up to larger and larger docid sets. This makes intersecting faster for most common index distributions.
Showing
Please register or sign in to comment