[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >
| In a message dated 10/4/2002 4:32:06 PM Central Daylight Time, lee@hidden.email writes: << When JCB started in the 50s it [self-segmenting] >> Ouch. Maybe in the 50's, but JCB was using computers up until he died, so he probably caught up with your watch (which sound pretty terrific -- unless you are factoring in time and size when you say powerful). << It seems to me that if present state-of-the art speech recognition software can perform word breaking about as well as a human using a reasonable amount of resources, then making that part of a language design is a waste of time. >> But even 2% error (and I have yet to see a program that lived up to that claim in normal running speech) can make for 100% errors when fed into a complete but rather stupid parser (viz. ours). And it is that damned parser that is Lojban's chief claim to be of interest to anyone -- despite people's ongoing efforts to subvert it. |