Program filter showing some bad results

ec-doris / kohesio-backend

APIs serving Kohesio's frontend

https://kohesio.ec.europa.eu

6 stars 2 forks source link

Program filter showing some bad results #250

Open raphdom opened 5 months ago

raphdom commented 5 months ago

We have a problem with the program filter, showing some bad interreg programs for Belgium, for example:

https://kohesio.development.ec.europa.eu/wikibase/facet/eu/programs?country=https://linkedopendata.eu/entity/Q8

DiaZork commented 5 months ago

By default the behavior of this api is that you get all program of a country, interreg included Then there is the interreg field that

if set to true return only interreg program (ie having the statement P579160 Q4554132)
if set to false return only not interreg program (ie not having the previous statement)

I can change this behavior if needed but you can also just call the API with a false interreg

raphdom commented 5 months ago

I know, sorry I was not so clear:

For Belgium apparently we have some Interreg results that are not meant to be there, like this:

2014TC16RFCB006 - Interreg V-A - Spain-France-Andorra (POCTEFA)

@madewild and @robertbrowncc can confirm?

DiaZork commented 5 months ago

Okay so i guess the question is Is it normal that Belgium is in Interreg V-A — Espagne-France-Andorre (POCTEFA) because it's a data problem in that case…

raphdom commented 5 months ago

yeah, maybe it is not a backend problem, feel free to transfer this to another repo, Thanks @DiaZork

madewild commented 5 months ago

I know, sorry I was not so clear:

For Belgium apparently we have some Interreg results that are not meant to be there, like this:

2014TC16RFCB006 - Interreg V-A - Spain-France-Andorra (POCTEFA)

@madewild and @robertbrowncc can confirm?

I think you meant @roberto-musmeci and not @robertbrowncc ;)

madewild commented 1 week ago

I removed Belgium from https://linkedopendata.eu/wiki/Item:Q2463550, let's see if it fixes the issue...

madewild commented 1 week ago

I also removed Belgium from a few others... Dunno how it got in there.

Dgojsic commented 1 week ago

Just seeing this now, in the case Raph mentioned (2014TC16RFCB006 - Interreg V-A - Spain-France-Andorra (POCTEFA), i looked at the source data, and there is indeed 1 beneficiary of 1 project located in Brussels. So the issue was most likely that for interreg projects (at least if they were fed from keep.eu, which doesn't have project locations) the beneficiary location was used instead. Probably also the case in the other instances

madewild commented 6 days ago

Ok, but so was I right to remove Belgium from these programmes?