Skip to content

POIs and POI categories

Like so many other people (for example Thomas here, but I vaguely remember a whole bunch of blog posts in the last few weeks) I am currently in the midst of trying to write my own FourSquare/Swarm replacement. I still use it a lot – mostly for my own amusement and record keeping. And while...

Dominik Schwind
Nov 11, 20262 min read

ike so many other people (for example Thomas here, but I vaguely remember a whole bunch of blog posts in the last few weeks) I am currently in the midst of trying to write my own FourSquare/Swarm replacement. I still use it a lot – mostly for my own amusement and record keeping. And while the “social” bits were a bit fun in the early days when it was still FourSquare I didn’t really use it like that any more in the last couple of years.

And yet, every time I get going, I get stuck on the POIs/venues. I tried using Overpass/OSM but somehow I’m not really warming up to the tag-based classification system. I started to use the iD Tagging Schema to try to at least get some kind of category based on the tags. And then I started to think about FourSquare’s categories. Their category tree has been under a rather permissive license ever since they open sources a lot of their places data so I figured I would want to use that, especially since I a) quite like it and b) have a huge archive of my old checkins with the category ids attached to the venue data. Alas – this way lies madness. I’m not going through a bit more than 1400 categories manually. Somehow matching them to iD presets and then matching them up to tag combination seems a fool’s errand as well.

And as far as I can tell, there isn’t anything out there where someone did it before. The only thing that I sort of found was this OpenAccess paper which is pretty LLM-heavy and didn’t really help me much, either.

So I guess in the end, I’ll end up using a more manual process, which is a bit of a bummer, especially since I was hoping to wrap all of this into a neat little library. Oh well.

Did you enjoy this article?

Recommend it — Standard Reader surfaces well-loved writing to more readers across the network.

Across the AtmosphereDiscussions