3 ms·
I actually worked on a service for a company in the past that could identify these types of scenarios and determine if they are fake. There were a number of sc
by eek2121 5y ago
I actually worked on a service for a company in the past that could identify these types of scenarios and determine if they are fake.
There were a number of scenarios to consider (there was a "probability score" involved), but the biggest one is: "How big is the plot of land at address XYZ. Is it in an urban, suburban, or rural area? Is the existing building a house, commercial building, or skyscraper? Does this unit buy products of this type in bulk?"
Easier than you think. Note that the score had hundreds of factors. These were the easiest to consider.
(In most states, you could pull public info and tell right away if an address has multiple legitimate units or not FYI)
- tyrfing 5y agoI'm curious how you handled addresses that couldn't be validated. From my perspective, that's a big blind spot of real-world implementations of this sort of stuff, and thus the condition under which most of it happens. Not quite "make up a unit number" (freight forwarder red flag, anyway) but pretty close.