You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The two queries have already different result lists when used without the location bias. We currently structure the query such that the location bias more or less only results in reordering of the search results without location bias, so that the closer result is never really considered.
lang=default gives so much more of a boost to 'Primekens Kopf' in 'Permecke' (both somewhat similar to 'q=primeke' that the other possible matches are discarded.
This needs some deeper structural changes to the ES query to get fixed. Possibly an early result boost for location bias and bbox.
See the following requests:
https://photon.komoot.io/api/?q=primeke&lang=default&lat=59.863704493020094&lon=17.639317368224624
https://photon.komoot.io/api/?q=primeke&lang=en&lat=59.863704493020094&lon=17.639317368224624
Both OSM nodes have only a "name" tag and no "name:en", so that I wouldn't expect a difference.
The text was updated successfully, but these errors were encountered: