Registering and logging in removes this ad.
Registering and logging in removes this ad.
no option for smaller vocab
Submitted by gardenerstouch on Thu, 09/27/2007 - 06:34.
I've read on a number ofr occasions that to improve the speed of speech recognition, that I should select the smaller of the two regular vocabularies when setting up a use account. I did indeed see this in DNS 8 but, I can't find where in 9.5 that could be done. Any suggestions?
When asking questions of
When asking questions of this type please let us know PRECISELY what NaturallySpeaking product you are using. Example: Standard 8, Preferred 9.5, Pro 7.3, KnowBrainer 2006 etc.. The additional information helps us to focus our response. DNS simply isn't enough information.
In Ver. 9 your only options are the Empty vocabulary or the bloated general large 150,000 word vocabulary. We will be releasing a medium sized 50,000 word vocabulary as part of our KnowBrainer 2006 and UniVoice utilities within the next few weeks but it WON'T be easy to utilize a 50,000 word vocabulary in DNS Preferred because of it’s limitations. It will be easy to do this in all the professional versions of NaturallySpeaking.
This question has seen a lot of discussion on the KnowBrainer Technical Support Forum
Lunis Orcutt - Developer of KnowBrainer (DNS Command Software)
Now Providing FREE (1st 5 min.) Tech Support 615-884-4558
A Nuance Gold Certified Endorsed Vendor
I've heard it said that the
I've heard it said that the average person uses only 4 to 5 thousand words in their verbal vocabulary. So, theoretically, I could use the Blank Vocab and import my own list of words, and I'd have a real whiz-bang speed increase? What are the caveats to that approach? Would I have to train each word individually?
Never mind, I found the
Never mind, I found the answer over at KnowBrainer. Thanks anyway!
gardenerstouch wrote: I've
I've heard it said that the average person uses only 4 to 5 thousand words in their verbal vocabulary. So, theoretically, I could use the Blank Vocab and import my own list of words, and I'd have a real whiz-bang speed increase? What are the caveats to that approach? Would I have to train each word individually?
Actually, the average person has a verbal vocabulary of between 25,000 and 30,000 words.
However, don't bank on blazing speed simply because you have a smaller vocabulary. That does not follow necessarily. Although smaller vocabularies will decrease the general response time to a certain extent, there are too many other variables that influence such. It is more likely that slower systems, such as single core Pentium 4's and the older dual core systems with less than 2 GB of RAM will experience a more dramatic increase in performance and accuracy using smaller vocabularies. However, Core2™ Duo and Core2™ Quad systems with two or more gigabytes of RAM generally do not perform any better with smaller vocabularies. I get instantaneous display of text and virtually 99.9% accuracy using the general large vocabulary with all 150,000 words with my Core2™ Quad running 4 GB of RAM.
In addition, you will likely find that creating your own vocabulary from a user set up as "Empty Vocabulary" is likely to be more trouble than it's worth. It isn't simply a matter of adding your own vocabulary list. The empty vocabulary has no vocabulary at all and creating your own vocabulary list requires that you have a thorough understanding of the necessary words (i.e., articles, nouns, pronouns, adjectives, verbs, etc.) that are required for normal every day standard speech/writing. I think you'll find this a very intimidating task, and, if done improperly, will likely result in poor accuracy, and more likely no accuracy at all (25% to 35%).
Chuck Runquist
Former Dragon NaturallySpeaking SDK & Senior Technical Solutions PM for DNS
If computers get too powerful, we can organize them into a committee - that will do them in. - Bradley's Bromide