search

osm-fr / osmose-backend

Part of osmose that runs the analysis, and send the results to the frontend.
GNU General Public License v3.0
92 stars 115 forks source link

Remove "Duplicated similar values" #779

Closed maro-21 closed 4 years ago

maro-21 commented 4 years ago

Can we remove item 3060 Duplicated similar values?

Looks like there are more false positives than something that needs to be corrected. In fact, I didn't find any error that needed improvement.

In my region there are thousands of errors like this:

I also don't see what's wrong with these ones:

Subitem "duplicated values" looks ok. It finds errors like:

But for "Duplicated similar values" I don't see any issues that need to be corrected.

http://osmose.openstreetmap.fr/pl/errors/?item=3060

frodrigo commented 4 years ago

Let's try first to improve it https://github.com/osm-fr/osmose-backend/blob/master/plugins/TagFix_DuplicateValue.py

frodrigo commented 4 years ago

With the first fix we already loose a large part of false positives.

frodrigo commented 4 years ago

http://osmose.openstreetmap.fr/en/errors/?item=3060&class=30601 http://osmose.openstreetmap.fr/en/errors/graph.png?item=3060&class=30601&start_date=2020-02

maro-21 commented 4 years ago

The graph looks great :) Thank you, good job!

Can we also add phone, fax, contact:phone and contact:fax, to the blacklist? And alt_name?

frodrigo commented 4 years ago

Ok for phone and fax.

But not sure about alt_name, wiki say to avoid multiple values in it.

maro-21 commented 4 years ago

Didn't know about it. Strange. What if there are more than one alternative names?

Looks like alt_name together with old_name dominates here Examples: Duplicated similar values alt_name=Chaykendy;Chaykhaneh;Chaikhana Duplicated similar values alt_name:be=Юзафоўка;Юзіфоўка;Юзяфаўка;Язэпоўка Duplicated similar values alt_name=Geydarabad;Haiderabad;Haidarabad;Hedarabad Duplicated similar values alt_name:en=Aghajaglil;Aghaj Owghlu;Aghaj Uqli;Aqajoli;Agadzhagly;Aghaj Oghlu

If the Wiki suggests to avoid such, what it suggests to do instead? If alt_name:en=Aghajaglil;Aghaj Owghlu;Aghaj Uqli;Aqajoli;Agadzhagly;Aghaj Oghlu is incorrect, how to correct it?

We should also blacklist these ones:

maro-21 commented 4 years ago

We should also exclude:

frodrigo commented 4 years ago

Updated, but low impact.

More to disable can be found here: http://osmose.openstreetmap.fr/en/errors/?item=3060&class=30601

frodrigo commented 4 years ago

Open new issue if required.