search

iptc / sport-schema

The next generation of sports data, based on IPTC’s SportsML and semantic web principles
15 stars 1 forks source link

Detailed modeling of (F1) auto racing #21

Open bobdc opened 3 years ago

bobdc commented 3 years ago

I don't know how much sport-model has done with auto racing, but I recently learned about an excellent set of Formula 1 data, so their relational model for drivers, tracks, races, and even pit stops and more (complete with instance data!) may be useful to your efforts.

You can see links to various aspects of it at the bottom of https://www.data-is-plural.com/archive/2021-08-25-edition/.

bquinn commented 3 years ago

Thanks @bobdc !

FYI SportsML, which our sport model (yet to be named!) is based on, does have a motor racing "plugin" - not specifically for F1 though. See http://iptc.org/std/SportsML/3.1/specification/sportsml-specific-motor-racing.xsd

I've clicked around the Ergast site a bit and it's an awesome resource, all free and mostly created by hand!

Here's their XML Schema: https://ergast.com/schemas/mrd-1.4.xsd

It's possible to download the entire dataset, and data re-use appears to be free (although using the API for commercial purposes is not), so we could have a go at transforming it to our model once it's complete. It would certainly be an interesting exercise.

And thanks for reminding me about Data is Plural, I get their emails but never get around to opening them :-)

pauljkelly commented 3 years ago

Thanks, I've bookmarked it. Probably the main modelling issue with be the vehicles which, like horses in horse-racing, have the same status as the driver/riders and require a different set of descriptive properties.

bobdc commented 3 years ago

I figured that its modeling of driver/team/track/race relationships would be handy in other forms of racing.

(I'll probably convert all the Ergast instance data to RDF so that I can query it in SPARQL, but that's me. Maybe I'll do it while watching the Italian GP this Sunday morning!)