Blogging’s been light lately because, incredibly, I am actually being productive in work-related tasks. I think that’s mainly because I am working at home, where I tend to get a lot more work done than at the office, library, or coffee shop.
I’m doing background reading for the Reconnect conference coming up at the end of next month, reading through Foundations of Statistical Natural Language Processing (“SNLP”) by Christopher Manning and Hinrich Schutze. It can be slow going at times, but I am finding this subject very interesting. It’s a combination of probability, statistics, computer science, and linguistics, all addressing problems in modeling human language effectively in an automated system. For example, if you scanned in a text using optical character recognition, and the sentence “I had a good time at the sho_” was produced by the computer — with the blank representing an error where the character was omitted, how might the computer fill that in automatically? There are several options — “show”, “shop”, “shod”, “shot”, etc. — but which one is most likely? That kind of thing, and more complicated problems besides.
I’m finding computational linguistics — of which SNLP is only a part — to be pretty intriguing. In many ways, coming to this area brings me full circle in my intellectual life. In high school, my least favorite subject was math, and my favorite subjects were always languages — English, and then the two years of Latin I had in the 11th and 12th grades. I went on to major in psychology for a brief time in college — it didn’t appear to me that double-majoring in English and Classics held a lot of promise for jobs later on — and got interested and did some research in cognitive psychology, especially language acquisition and problem solving, and taking lots of stats. Then I switched to math. Now here I am, ten years into my math professor career, studying mathematical methods such as statistics applied to the study of language.
I’ve also blogged previously about other stuff I’m doing — the 3+2 engineering program, screencasting, retooling the mastery exams for Calculus Preparation, and prepping the textbook-free version of Modern Algebra I’m teaching this fall. I will have more detailed posts on these soon.