This week, I’ve been busy with drawing pictures for a report, and trying to catch up with my coursework. The latter is a pain for many reasons. I missed a few classes and exercises due to other activities and in natural sciences, that is BAD. You need to learn all the basics in the course if you want to succeed; and it’s better if you learn them well, because you’ll be needing them later.
Because I wanted to jump into the data finally, I considered skipping the course altogether, but then my professor urged me to push it through. >.< natural language processing seems to be a combination of probability science, language theory, algorithms and programming. All are things I need, but while I recognize all these fields from the past, it is a tough set at least for me after all these years! (Luckily I have refreshed my stats and programming lately… but the Chomsky formal language part I’ve last seen more than 10 years ago.) I’m not yet sure how exactly to use these methods in social sciences analysis, but I managed to find one workshop proceedings document from 2014. On the other hand, selecting algorithms is made easier when the adjunct professor next to you has actually worked and developed them! And the pattern finding is appealing, it reminds me of bioinformatics I studied a bit back in the day.
While studying, I’m listening to Russian The Five (Polovtsian dances then Pictures at an Exhibition). Self-trained, provincial, promoting the Slavic culture instead of a Westernised version of Russia? Taken together as a team, they’re just as good as the more cosmopolitan-styled loner Tchaikovsky.