[Photon] location bias

[Sarah Hoffmann]

Hi,

On Mon, Jan 15, 2018 at 09:00:43AM +0100, Peter wrote:
> Hi,
> 
> When upgrading from the old to the new geocoding I can see a good
> average speed improvement overall.
> 
> One bigger headache I have is that the location bias previously wasn't
> that well working and now is very strict. Here is an example
> illustrating the necessary different strengths. Always assume I use
> "berlin" as location bias (e.g. from a guessed location via IP address).
> Now if I use the location bias and
> 
> 1. I type "railway station" (or "bahnhof") then I want stations of this
> city as first result, but
> 
> 2. if I type another city like Paris I usually want a mix of "Berlin,
> Pariser Platz" and results from the city "Paris".

These two searches are fundamentally different. One is a POI search
in your surroundings, the other is a search for a name. You cannot
mangle them into one and hope to get good results. If you want to
support 1), then you have to add some form of keyword detection for
POI terms and switch to a completely different kind of search. The
current strong location bias is not enough.

> Maybe I have to mix the results from two requests (one with the location
> bias, one without) on the client side? What do you think?

No, this won't work either. Applying each metric separately is not
the same as applying them all at the same time and balancing them
out carefully. The metrics here being location bias, the importance
of the place and how well the search string matches.

Location bias is something that in my experience needs to be added
very carefully. The importance is much more useful. For one thing,
distance searches are actually quite rare outside POI searches.
But more importantly, if the ranking goes wrong, it is almost
impossible for the user to correct a location-biased search
(try getting the real Paris' with the current photon version)
while with an importance-biased search, bad results can almost
always be corrected with being more specific (searching for
'Paris, France' for example).

Sarah