|
Top-level
(2/3) ...an info pattern like "traffic lights in Berlin that are located no more than 100m from a metro station and no more than 50m from a pharmacy" – because you have this image of a political rally you want to verify. *Spot* will then show you the matching results in a nice interface. No coding needed, just plain English. Now in order to come up with better *Spot* results, we need people who know the #OSM tag #database really well – and can help us navigate it in an efficient way. 16 comments
  @dw_innovation hi! Incidentally I gave two talks at CCC, one about advanced #OSM usage and one about #OSINT, which also features exactly what you're doing with this tool (using an example from one of Julia's quiztime posts ;) I regularly do this kind of research using overpass. This tool sounds great! Are you aware there's a very similar tool from @Bellingcat? https://github.com/bellingcat/osm-search 1/2 @dw_innovation The JOSM and iD presets are a good source for commong language to tag mapping. The Name Suggestion Index might also help. Feel free to reach out via Mastodon or Email. You can find my address and the talks on my website. Really looking forward to this tool! 2/2 @jomo @dw_innovation @Bellingcat Can you link your talks? I would be interested. Some issues I could think of here is the use of ChatGPT, you would probably want this a little more in-house. Also to be dependable for OSINT, you would need some sort of fuzzy matching, where not all features are a must or could deviate from what you described. Lastly some sort of inclusion of historical changesets may be great, since places change all the time :) Many thanks for your favs and boosts and most of all: your helpful comments, links, and tips! We'll check out all of them. And we'll get back to individual users if there are more specific questions. Cheers! @dw_innovation this looks like a fantastic use of ml... assist in a task and allow the user to verify the results. I can see this being an incredibly useful tool if you can get it to where it needs to be.  @dw_innovation You can start looking at the #OpenStreetMap editor software presets. There 2 main ones, the iD one ( https://github.com/openstreetmap/id-tagging-schema/ ) and the JOSM one https://josm.openstreetmap.de/wiki/Presets The database of brands→OSM tags (the “Name Suggestion Index”) is also informative https://wiki.openstreetmap.org/wiki/Name_Suggestion_Index read more about preset: https://wiki.openstreetmap.org/wiki/Preset  @amapanda @dw_innovation I would naturally argue that https://github.com/simonpoole/beautified-JOSM-preset is superior to both 😎 But more important, give https://taginfo.openstreetmap.org/ a look. There are many considerations to be made for matching (something that modern editing apps need to do), for example handling multiple "real world objects" that are tagged on a single geometry and many more. @dw_innovation Do you have concrete questions? The question is quite broad here. @dw_innovation nice, aber im wesentlichen fast umständlicher als dieses Tool OH: gegen Falschinformationen wird mittelbar eventuel ein neues scoring system der Menschen untereinander helfen. Bei gleichzeitiger individueller Gewichtung. Da gleichzeitig ausgewogen und komplex ist. Eine Art Web of Trust 2.0. einiges an Hirnschmalz nötig. Tivialste Form wäre eine art "probabal/wahrscheinlich" Button. @dw_innovation I search for the tags normally on the OSM wiki which points also to alternatives: https://wiki.openstreetmap.org/wiki/Main_Page Maybe yo can integrate in *Spot* the Wiki |
Gregory's Server
(3/3) So are there any #OSM experts out there, per chance?
We'd love to hear from you! ❤️
For more info, check out our #paper on this: https://arxiv.org/abs/2311.08093
And the #code repository on Github: https://github.com/dw-innovation/kid2-spot