Hi All,
Can you please explain how ranking is done in fuzzy AND normal full-text search?
Here using wikipedia topics as test, and still have very odd results:
start n=node:topic("name:'DNA'~0") return n limit 6;
start n=node:topic("name:'DNA'~0.3") return n limit 6;
return same Sakartvelos Gogona Skautebis Asociacia 'Dia'
as first result;
changing fuzzy to ~0.8 gives no results.
start n=node:topic("name:'united states'~0.3") return n limit 6;
gives no result for any fuzzy parameter.
start n=node:topic('name:united states') return n limit 6;
gives:
List of United States National Historic Landmarks in United States commonwealths and territories, as...
as first result, instead of 'united states' as expected (which is present in my db).
How does fuzzy search and normal START search in legacy index work?
I need to hit more pertinent results and do a little traversal in the same query.