I am surprised that Dragon NaturallySpeaking is so bad. It can keep going for a little while, but it always loses it after a bit of dictation. For instance, today after about 40 minutes of dictation it stopped without warning due to lack of memory or something. WTF? I have 6-8 GB of memory and a basically empty 1 TB hard drive. How does anything run out of memory?
I want the system that Commander Adama used on the old Battlestar!
So far research in machine learning (for me anyway) is really writing utility programs for text file processing. :-/
I know what stemming is, it is making a computer recognize that ‘burp’, ‘burping’, and ‘burped’ are all really the same word ‘burp-’. But what the heck is lemmatization? (Off to Wikipedia!) :-)
So today I am spending the day learning about ways to make computers recognize that ‘sexy’ and ‘sexiness’ are the same concept (but ‘sex’ is not, so there may be a problem there!). That is my afternoon. Doing stuff like that. And python.
Doing stuff like that in python.