Saturday: Surprisingly, natural language processing became something I thought about this evening.

Even without actual language processing, I was able to go from concept to initial parsed data in less than an hour.  Regexps aren't an optimal strategy, though.  I talk a lot about pizza, but without parsing the language, you wouldn't know that this post was actually about sushi.

I could have it do more detailed parsing, so that sushi contains a class of descriptors, so me mentioning the miso soup and the unagi adds weight to that topic.

Side note: salmon belly wasn't as good as I'd hoped.  It's filled with burning oil, until it cools off, then it kind of just isn't very good.

Okonomiyaki special.  A disappointment, given that one of the most delicious things I've ever eaten was an okonomiyaki.  The tako inside was rubbery, and the flavors didn't blend well.  The katsuobushi didn't dance, either, which was unfortunate.

So ahi and tamago to make up for it.

Unfortunately there was a gravity related accident.  :(

• If I hadn't gotten lazy and stopped, using this language parser for perl would have been the next step.  Also if I wasn't lazy, I would have used some of the html information in my blog dump instead of just stripping it all out.
• I assume the answer is "some stupid laws," but can't California just institute a scaling water price?  Target a usage quantity, determine the per capita value, and then ramp prices above that value up until the actual usage matches the target.  Rich people may be rich, but there's a point where you can force them to change their actions.  If that next gallon they use costs $1000, they're probably going to be less of a dick about it.
• I didn't remember this scene, and that's so disappointing because it's brilliant.
• Dog science.
• Squirrel.
• Marnie.

Also, thinking about it here at the end after doing the links, it'd be nice if there was a easy way to auto-reblog all the things from tumblr that I link to.  Just something parsing my blog data made me think about.