Skip to content
This repository has been archived by the owner on Mar 15, 2024. It is now read-only.

Common compare function for checking feature names #133

Open
bkowshik opened this issue Mar 30, 2017 · 8 comments
Open

Common compare function for checking feature names #133

bkowshik opened this issue Mar 30, 2017 · 8 comments

Comments

@bkowshik
Copy link
Contributor

Adding names and name translations happen pretty often on OpenStreetMap. Yay!!! Quite of few compare functions currently flag any edits to features that satisfy a particular criteria. Ex:

  • Edited a major lake flags any edits to features in the list of major lakes.
  • Edited an osm landmark flags any edits to a custom list of landmarks on OpenStreetMap.
  • Edited a name tagflags any edits to feature name.
  • Edited an old monument flags any edit to monuments on OpenStreetMap.

For a large majority of cases, the new names or translations added are 👍

screen shot 2017-03-30 at 2 49 09 pm

How about we move the name checks to a common profanity compare function and flag any other kind of edit from the ^ compare functions? The profanity compare function in turn will focus just on identifying problems with values of name:*.


cc: @amishas157 @geohacker

@bkowshik
Copy link
Contributor Author

A couple of compare functions work with feature names. Ex:

We should 👀 if there are opportunities to merge them into this common compare function.

@bkowshik
Copy link
Contributor Author

From @nikhilprabhakar's post here: #110 (comment)

case where the user named the feature after the user's name

screen shot 2017-03-31 at 11 32 43 am

User MattForce could have potentially named the lake after himself as Lake Matt

Should the name compare function do this check as well?

@bkowshik
Copy link
Contributor Author

bkowshik commented Apr 2, 2017

Changing the order of words is an interesting scenario. Ex:

The name of the feature was modified from Воскресенская церковь (Resurrection Church) to Церковь Воскресения Христова (Church of the Resurrection).

@bkowshik
Copy link
Contributor Author

bkowshik commented Apr 2, 2017

Minor name modifications can be ignored for a while. Ex:

The features name was modified from Thun - Hohenštejnský palác to Thun-Hohenštejnský palác, which is a couple of blank spaces was removed.

Palác Thurn-Taxisů modified to palác Thurn-Taxisů

The name was modified from St. Joseph Hospital to Saint Joseph Hospital

@bkowshik
Copy link
Contributor Author

bkowshik commented Apr 2, 2017

There are times when the case of the name (uppercase / lowercase) does not matter much. We can ignore these for a while. Ex:

The name was modified from Aldi to ALDI.

@bkowshik
Copy link
Contributor Author

bkowshik commented Apr 2, 2017

Thank you for your feedback @bsrinivasa, 😃

@bkowshik
Copy link
Contributor Author

bkowshik commented Apr 5, 2017

For the first iteration of this compare function, let's do the following checks:

  • If feature has a Wikidata tag, does the feature's name match the name on Wikidata
  • Is there any profanity in the feature name
  • Was this is a major name change or a minor one using [Levenshtein distance](https://en.wikipedia.org/wiki/Levenshtein_distance

cc: @amishas157

@srividyacb
Copy link
Contributor

Everyday a lot of features get names and many features names gets deleted. Found an edit where the feature tag and name string is similar

Examples :

This kind of name changes are 100% incorrect and we need to flag it as part of our compare function. The name combination can look like :

  • Feature tag + name string are similar (with some case changes)
  • name tag with any of the standard tags

Would like some suggestions on this.

cc @bkowshik @amishas157

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants