pscheid92
commented
4 years ago Unfortunately, I don't see a description of the data processing pipeline or the processing in general. Is there automated ingestion or a cleansing step one could use to map to some kind of place id?
It seems odd that we, as a worldwide consortium of nerds and geeks, can't figure out a better way to hash out locations.
That said, as we know, the researchers uploading raw data do not have time for this, so it falls to the community to find a way to make it work.
In theory, we would be able to map out:
Between FIPS, ISO3166 and it's variants, GENC (example), or (Gods forgive me for suggesting, the very robust Google Places) see image below and https://jsfiddle.net/ocL8en97/2/#&togetherjs=O8eOaSM9hL - search for a town on the lists and it will have a PlaceID.
By using a standards based mapping protocol, it would be significantly easier to trace community spread, as well as to alert the public as to their local situations.
Using only the lowest denominator (town/city/village), the upstream identifiers are available, as seen in the GENC example. So we easily see 'Luxembourg>Wallonia>Belgium>Europe' as the geodata string.
In addition, one could theoretically then tie into municipal databases such as health departments, cemeteries, etc. to track demographics. Couple that with population databases and suddenly it could be a very robust method of prediction as well.
Since we are getting Lat/Long data, it would seem a fairly logical jump to link those to a location ID of a standardized placeID.
For instance, 47.7511 -120.7401 leads us to Leavenworth/WA/US.
Frankly I just do not a)trust the governing bodies to put in the work or b)to get it right.
Thoughts, drawbacks, smackdowns?