Is there any public information on the underlying sources for Safegraph Places data? I cannot find any mention in the online manual. (Not looking for a list of sources, just trying to get a general sense of what types of sources they use — yellow pages? Openstreetmaps? Yelp?)
Hi @Cody_Cook_Stanford_GSB, from the Safegraph website, they:
How We Create Our POI Database:
INGEST
We onboard data from thousands of diverse sources. We compare, de-dupe, cross-reference, and discard bad data.
DRAW
Our building footprints are derived from satellite imagery and supplemented with hand-drawn polygons.
MERGE
We use machine learning and human feedback to associate POI business listing info with building footprints.
CLASSIFY
We algorithmically classify brands and denote spatial hierarchy. POIs (like restaurants) can exist within other POIs (like airports).
VERIFY
We leverage unique truth sets to continually improve the accuracy of our datasets.
Basically, they crawl the internet in thousands of places and ways to create this massive dataset.
I saw that. It’s quite vague so I was hoping for something more concrete.