[Talk-us] GNIS Import Done

[Matthew Schneider]

> If you know that Zavala School and Zavala Elementary School are the 
> same thing, then you should delete the "less correct" node (along with 
> its gnis:feature_id). That would eventually push a "delete the less 
> correct data point" change into GNIS, which is a good thing I think.
I hope soon to start some analysis based on the names of GNIS nodes. 
Initially to attempt to recategorize secondary medical facilities such 
as clinics from true hospitals. I will note that I'm not sure what to do 
with secondary facilities once I have sorted them out; amenity = 
hospital really doesn't see appropriate; all other proposed tags seem to 
have been voted down. We might drop them, or assign them a non-standard 
tag for later reclassification.

On top of that, it might be appropriate to perform some sort of name 
based analysis on nodes with the same class and within a distance of N 
units from each other. For items like schools, stripping common words 
from the names ("Elementary", "High", "Junior", "Middle", "School") and 
comparing the remaining values one word at a time might provide a quick 
list of redundant nodes for review.

If anyone is interested in participating with this effort on a more 
national level, I'd welcome the help.