The difference in association patterns between "Fellows of the Water Buffalo Lodge" and "Drug dealers and their clients" is probably not one that is trivially s

The difference in association patterns between "Fellows of the Water Buffalo Lodge" and "Drug dealers and their clients" is probably not one that is trivially solvable by algorithm. There's plenty of good reasons that someone could want to make a bunch of short stops along a circuitous route - Not the least of which is commerce of precisely the same nature, but with legal substances. I know somebody who bakes cookies, sells them online through facebook, and takes uber to make her deliveries. It's precisely the same scenario, but nothing illicit is going on. How would an algorithm distinguish?