from https://simulationhockey.com/showthread....pid3573536 +2

6
Agate would definitely make those like 3 hour videos of her just doing work and projects on her remote farm in the middle of the most beautiful place you’ll never get within 100kms of. One of those ones where the first video is her cutting an hour long driveway from the nearest dirt path to whatever clearing on her property she decided would be base camp. And she’d insist on trying to use the most labour-inducing tools possible, except when the neighbours show up with one of those Indiana Jones Jungle Cutters, and save her years of clearing. Would be great functional training for hockey though! +2

11
So actually, I already responded to a Chirp this week, talking about how someone should build a Natural Language Processing project built on our existing Chirp Data and then ask it questions about life and hockey. I think it would be complete and utter garbage, but I still want to do it, especially considering I’m taking Natural Language Processing classes this semester and next. I’m already working on some other Natural Language Processing project ideas for other leagues, and so once I have the foundations down, I think it would be pretty easy to crawl the Chirp pages to get all of the Chirps and give them some structure. In fact I already conceptually know how I would do the project, I just haven’t seen the implementation yet to go off of. I’ve built predictive models, I’ve been deep learning models, and I’ve done all the text pre-processing like a million times now, but the combination is something novel to me. Could I stop being lazy and figure it out myself? Sure, but then what am I paying college for? These are exactly the kinds of projects that I want to use to get ready to do actual portfolio projects, to build those multi-use modules that will save me a ton of time down the road! +4