daveywoods

I'm open to super terrific ideas that I might not be aware of. I'm the world's least expert website scraper / data miner / white hat hacker who is actually running a scraping function like this. Some of the folks here for sure would be more adept than I am at seeking benign access to the full dataset. I'm scraping the message board front page only, using a python Lambda, through scraperapi. There is no API or RSS feed available to me, per the board admin. There seems to be a chance that could be available in the future.

Here's a neat weird thing I tried for fun. This is an AI generated "temperature" of the TSL football message board, and other assorted insights culled out of posts here. It's all fully automated. I run it on an Ohio State and aTm board also to compare and see if the thing works there too. Check it out. I don't know what I'm doing with this thing, I'm not trying to get clicks and whatnot. Ideally I'd like to get better input data so the thing can be smarter. For the fellow nerds: There's not as much input data as I'd like. Scraping TSL, even with Will's generous help, doesn't work as well as I'd like. I'm not clicking into individual messages, I'm using subject lines only on one scrape of the main index. Yes this is all vibe coded py and tf on AWS Lambda and DynamoDB.

I'm working on a project that is too nerdly to explain much. I just need your help. What is YOUR current Hokie hype vibe from 0 to 10? After the season we'd all probably be a 1.5 But there's this new coach And this recruiting class And hope springs eternal I'm at about a 4.0. I'm getting a 5.0 vibe from the people I talk to. Which is odd. Is CJF creating that much positivity? What is YOUR current vibe for Hokie football on a 0-10 scale? For science. Seriously I'm working on an project I need to know the fan vibe please halp. -Dave

I'm working on a TSL AI, just for fun. I ran this by Will, he said OK. I'm trying to build a sentiment analysis tool to evaluate TSL message board posts for Positive, Negative, or Neutral. If it works, I'll give it to TSL (or I'll run it for TSL). It will be interesting to see how message board sentiment trends through the season. This would be run on the bulk of the board, not labeling individual messages. Something like "this week the message board is 63% more positive than last week". This is achieved by grading individual messages, but only aggregate data would be reported. But assessment of positive or negative is obviously a subjective evaluation, which is why I'm here asking for help. I'm looking for people to "grade" the AI training data. That means looking at a few hundred rows of data and marking them as positive, negative, or neutral. The way AI works, if we give it human input on some data, it will be able to go ahead on an automated basis. If I train all the data myself, the sentiment analysis will be skewed by my bias. And anyway I could use help looking at all this data. If you're interested, email me and I'll send you 300 rows of messages to look at. You can do as many or as few as you like. This is simply looking at the data I have pulled off the message board for subject line and message, and grading that message as positive, negative, or neutral. Please help if this absurd nerdery interests you. my email address is my username at gmail dot com. Also I should mention -- Pretty good chance this never works, but let's try. It's neat.

I was always happy with Stiney, because Beamer was happy with Stiney, and I knew less about football and Stiney than Beamer knew about football and Stiney. I was also happy with perennial 10-win seasons. Those days sure were miserable around here with all the complaining though. Glad we're done with those days. [eyeroll]