In article <eBXbFelyJHA.6100@TK2MSFTNGP03.phx.gbl>, Wendy
<none@nospam.invalid> wrote:
[color=blue][color=green]
> > Don't allow this junk to overwhelm you, just delve into it
> > a bit at a time, when you are in the mood. :)[/color]
>
> That is what I am doing at the moment. Fascinating stuff
> and a great program by the way you portray it.
> Thank you for the advice.[/color]

Just keep in mind that I am biased towards Dragon, because
I enjoy the added flexibility of that app.

Dragon can be very frustrating at times, it is not a very
"intuitive" speech app, in my opinion.

In the hands of someone who already has speech recognition
experience, it way out-performs Vista Speech, as shown by
the very large community of professional Dragon app users,
as compared to the much smaller community of
Vista professional speech users.

In todays world, if a student is taught Dragon, he/she will
have no problem finding a job in a field that uses speech apps.

If a student is taught the easier Vista Speech, he/she will find
their opportunities much more limited.

That fact is not liable to change in the next decade, much as we
all (including me) - prefer the easier-to-learn and
more intuitive Vista Speech.

My chances of convincing a hard-headed hospital administrator
to junk Dragon and adopt Vista Speech are zero, despite
Microsofts heroic efforts to break into the
lucrative medical market.



[color=blue][color=green]
> > You might keep in the back of your mind that Dragon,
> > at the present time, has an upgrade sale going, until
> > only June 30, 2009.
> >
> > $300 (US) gets you an upgrade to the $900 "Pro"
> > version of Dragon 10.1
> >
> > That is a saving of $400 (USA)[/color]
>
> That is a lot of money for me. I should at least learn
> what I have at the moment do you not agree
> before I move on to updating it?[/color]

Yes indeed, I had second thoughts about alerting you to
that opportunity, after I posted it.


[color=blue][color=green]
> > My two dollar tutu is too tight to wear,
> > my other tutu was 2.2
> >
> > Why? Because I trained it that way. (in Dragon)[/color]
>
> May I ask how you as you say, "train it" to make it do
> what you have done above?[/color]

I was afraid you might ask that ;-)

The easiest way to explain the so-called "training" is
with this example:

If you dictate this sentence:

My list of words is to two 2 tutu too 22



....then this is what Dragon hears you speak:

My list of words is to to to to to to to to

(spoken at a normal rate, with no pauses
between the various "to" sounds)

Remember this is with highly flexible Dragon,
I did not try it with Vista Speech.

Took me one minute to train it.

Takes a lot longer to explain than actually doing it.
******************************

1) In the "DragonBar" menu strip at the top of my
screen, I selected "Words" menu.

2) Selected "View/Edit"

3) In the "Written Form" box, I typed:

to two 2 tutu too 22

4) In the "Spoken Form" box I typed: 22222222

(only because 22222222 was what I usually get
whenever I speak "to to to to to to to to")

5) Pressed the "Train" button, window pops open,
in the window in big letters is 22222222

Did not let that deter me, I pressed some button to
allow me to "Train" it, I forget the name of the button,
it might have been named "Go".

I promptly spoke: to to to to to to to to at a normal
rapid pace.

6) Closed the training window.

That is it, supposedly now that I slapped my Dragon
around a bit, everything is "good to go".


Let's find out, I dictated this:

My list of words is to to to to to to to to


Result:

My list of words is to two 2 tutu too 22


Just for variety, I dictated:

Wendy's list of words is quote to to to to to to to to quote


Onscreen text result:

Wendy's list of words is "to two 2 tutu too 22"


Perhaps someone else here could volunteer info
about whether or not Vista Speech is up to doing
the same one-minute training.

Shame on you Wendy, asking lazy old me tough
questions like that ;-) ;-) ;-)

Just kidding, of course.

Mark-

In article <010520091514050211%none-of@your-business.invalid>, Mark
Conrad <none-of@your-business.invalid> wrote:
[color=blue][color=green]
> > May I ask how you as you say, "train it" to make it do
> > what you have done above?[/color]
>
> I was afraid you might ask that ;-)
>
> The easiest way to explain the so-called "training" is
> with this example:
>
> If you dictate this sentence:
>
> My list of words is to two 2 tutu too 22[/color]


Whoops, I better add that the "sentence" above is only
for demonstration purposes, the sentence itself does not
make any sense.

The technique should only be used when normal correction
methods do not work for you, Wendy.


Like most good things that look too good to be true, this
High Powered correction technique has its drawbacks.

For example, that list of words, "to two 2 tutu too 22"

Let's say you now need a slightly different order of the
same "words", such as "tutu 22 too two to 2".

Will not work, Dragon will slavishly print the first list
that you trained it on.

Furthermore, if you train Dragon on *BOTH* lists, then
Dragon will get confused so you will get one list sometimes,
and the other list at other times, no way to predict which
list will pop up.


- - - - - - - -
Speech recognition is far from being an exact science,
it could be decades or possibly never, for a computer to
be as good as the human brain in converting speech to text.


Despite that, speech recognition is useful in many cases,
particularly when the original audio can be saved with the
text, as is the case in the "Pro" versions of Dragon.

That means a highly skilled scientist could delegate the
correction step to an assistant.

If that highly skilled scientist was working on a solution
to a "Swine Flu" pandemic, thousands of lives could be
saved by a speedier solution.



Another example -
----------

A PhD working on getting us to Mars could use his
valuable time to dictate everything associated with his
project, at normal speaking speeds of say 150 wpm.

An assistant could correct the PhD's text, so the scientist
could be extremely productive, by not needing to waste time
correcting his own text.

Mark-

Mark Conrad wrote:
[color=blue][color=green]
>> If you dictate this sentence:
>>
>> My list of words is to two 2 tutu too 22[/color]
>
>
> Whoops, I better add that the "sentence" above is only
> for demonstration purposes, the sentence itself does not
> make any sense.
>
> The technique should only be used when normal correction
> methods do not work for you, Wendy.
>
>
> Like most good things that look too good to be true, this
> High Powered correction technique has its drawbacks.
>
> For example, that list of the the go up 12 words, "to two 2 tutu too 22"
>
> Let's say you now need a slightly different order of the
> same "words", such as "tutu 22 too two to 2".[/color]
[color=blue]
> Will not work, Dragon will slavishly print the first list
> that you trained it on.[/color]

I sort of get what you're saying Mark but being a beginner should I not
think of something that will be useful to me similar to your comparison and
then put that into my Dragon vocabulary?

Like you say, your comparison does not make any sense and I would not
probably use it in normal dictation but I understand the point you are
making so I should think of something that I will most likely use then train
it for that purpose? As you say, I do not want Dragon to be confused :-) I
would want it to be the best it can and mean something that I would most
probably use and therefore a useful piece of vocabulary within Dragon.

I will think of something and try it and report back, in the meantime,
perhaps you could think of a similar comparison with words that will
probably be used in dictation?

Now then, here the homework begins! I have to think of some words that I
will probably use! I see what you mean, delving into this seems to be a
bottomless pit. What have I let myself in for? :-))

Wendy

"Mark Conrad" <none-of@your-business.invalid> wrote in message
news:010520091731407686%none-of@your-business.invalid...[color=blue]
> In article <010520091514050211%none-of@your-business.invalid>, Mark
> Conrad <none-of@your-business.invalid> wrote:
>[color=green][color=darkred]
>> > May I ask how you as you say, "train it" to make it do
>> > what you have done above?[/color]
>>
>> I was afraid you might ask that ;-)
>>
>> The easiest way to explain the so-called "training" is
>> with this example:
>>
>> If you dictate this sentence:
>>
>> My list of words is to two 2 tutu too 22[/color]
>
>
> Whoops, I better add that the "sentence" above is only
> for demonstration purposes, the sentence itself does not
> make any sense.
>
> The technique should only be used when normal correction
> methods do not work for you, Wendy.
>
>
> Like most good things that look too good to be true, this
> High Powered correction technique has its drawbacks.
>
> For example, that list of words, "to two 2 tutu too 22"
>
> Let's say you now need a slightly different order of the
> same "words", such as "tutu 22 too two to 2".
>
> Will not work, Dragon will slavishly print the first list
> that you trained it on.
>
> Furthermore, if you train Dragon on *BOTH* lists, then
> Dragon will get confused so you will get one list sometimes,
> and the other list at other times, no way to predict which
> list will pop up.
>
>
> - - - - - - - -
> Speech recognition is far from being an exact science,
> it could be decades or possibly never, for a computer to
> be as good as the human brain in converting speech to text.
>
>
> Despite that, speech recognition is useful in many cases,
> particularly when the original audio can be saved with the
> text, as is the case in the "Pro" versions of Dragon.
>
> That means a highly skilled scientist could delegate the
> correction step to an assistant.
>
> If that highly skilled scientist was working on a solution
> to a "Swine Flu" pandemic, thousands of lives could be
> saved by a speedier solution.
>
>
>
> Another example -
> ----------
>
> A PhD working on getting us to Mars could use his
> valuable time to dictate everything associated with his
> project, at normal speaking speeds of say 150 wpm.
>
> An assistant could correct the PhD's text, so the scientist
> could be extremely productive, by not needing to waste time
> correcting his own text.
>
> Mark-[/color]

"Wendy" wrote:
[color=blue]
> I sort of get what you're saying Mark but being a beginner should I not
> think of something that will be useful to me similar to your comparison and
> then put that into my Dragon vocabulary?[/color]

It is probably very helpful if you do two things with Dragon Naturally
Speaking. First of all, go to Tools, Accuracy Center, and feed several of
your typical documents from the add documents feature. Next, try reading
those documents while and unseating carefully and speaking in phrases. You
should get very high accuracy. Also, the practice will acclimate you to how
the Dragon wants to be heard.

Marty Markoe, eMicrophones, Inc.

>"www.MyMSSpeech.com" <wwwMyMSSpeechcom@discussions.microsoft.com> wrote in[color=blue]
>message news:39E93C33->766E-4C97-8089-9A816BF71DA5@microsoft.com...[/color]
[color=blue]
> It is probably very helpful if you do two things with Dragon Naturally
> Speaking. First of all, go to Tools, Accuracy Center, and feed several of
> your typical documents from the add documents feature. Next, try reading
> those documents while and unseating carefully and speaking in phrases.
> You
> should get very high accuracy. Also, the practice will acclimate you to
> how
> the Dragon wants to be heard.
>
> Marty Markoe, eMicrophones, Inc.[/color]

Oh hi Marty, thanks for the tips.

I am still battling with windows speech recognition as well and the WSR tool
kit looks like a handy thing to have so I am keeping all my options open at
the moment.

Wendy

In article <Odbc7N0yJHA.5032@TK2MSFTNGP05.phx.gbl>, Wendy
<none@nospam.invalid> wrote:
[color=blue]
> I sort of get what you're saying Mark but being a beginner should I not
> think of something that will be useful to me similar to your comparison
> and then put that into my Dragon vocabulary?
>
> Like you say, your comparison does not make any sense and I would not
> probably use it in normal dictation but I understand the point you are
> making...[/color]

Wendy, your points are well taken, I should have used a
better example, a "real world" example that you might
actually find useful.


[color=blue]
> I will think of something and try it and report back, in the meantime,
> perhaps you could think of a similar comparison with words that will
> probably be used in dictation?[/color]

Okay, will do. Good grief, now I am in the same pickle as you are,
now I will have to think of a valid example. ;-)



Here is my original bad example:[color=blue]
> For example, that list of words, "to two 2 tutu too 22"[/color]

Okay, example junked, thank goodness.

In fact, I will steer clear of using homonyms altogether
with my next "good" example


Hmm, lemme see, what to use instead.

Well, I do not know what grade you teach in your job, so I will
pretend I am you, and am teaching a bunch of graduate students
about Einstein's problem understanding another genius of his
era, a scientist named Paul Dirack, the man who predicted
the existence of anti-matter.

Okay, "Dirack" is a tough name for Dragon to handle, and I know
that I will be constantly using that name while I am preparing
my next-day lessons for my graduate students, using my
handy-dandy Dragon speech app to prepare my lesson.

I dictate "Paul Dirack" just for kicks, to see what Dragon
will print. Dragon prints "pole direct". Bad Dragon !!!

All right, I spent my next 60 seconds training Dragon, using
the steps in my prior post.

Briefly those steps again are:

(start from "Word" menu, top of Dragon's screen)

Words > View/Edit > WrittenForm > Paul Dirack[color=blue]
> SpokenForm > pole direct > Train[/color]

(now speak "Paul Dirack" into
the training window and click Done)



Please do not be overwhelmed by the steps, believe me
you will get used to doing that 60-second training drill.


Now you can prepare you lesson, confident that every
time you dictate "Paul Dirack" that Dragon will have
no trouble whatever recognizing it.


[color=blue]
> Now then, here the homework begins! I have to think
> of some words that I will probably use![/color]

Not to worry, just start rabbling away into Dragon,
sooner or later it will make a mistake.

If it makes that same mistake over and over, it is
a candidate for your 60-second training drill.


[color=blue]
> I see what you mean, delving into this seems to be a
> bottomless pit. What have I let myself in for? :-))[/color]

Naw, I have confidence that you will be an expert in days.

In my case, I am 99% senile at 80 years old.

Besides that, my false teeth fall out and I wind up
gumming the microphone.

None of my synapses have snapped for years.



Now, for a future lesson, after you get your smarts,
we will delve into *WHY* I was showing you the
"hard way" to train Dragon.

In other words, why did I not just parrot the party line,
showing you the "accepted way" of handling training?

Here is why; dictate the name "Saul Kripke" into
your Dragon, then try to train it, using conventional
methods like saying "Correct That" and then saying
"Train". Chances are that Dragon will never learn.

Saul Kripke was a famous logician.

Just for beginners, his last name is an ancient Egyptian
name for a female. I could not find how it is pronounced
anywhere on Google.

I pronounce it "cripe" - am probably wrong.

Mark-

In article <030520090027215502%none-of@your-business.invalid>, Mark
Conrad <none-of@your-business.invalid> wrote:
[color=blue]
> Now, for a future lesson, after you get your smarts,
> we will delve into *WHY* I was showing you the
> "hard way" to train Dragon.[/color]

Whoops, better clear that up.

Heck, I use the standard "recommended" way of training
Dragon, most of the time it works great.

There is a lot to be said for keeping things simple.

The convoluted way of training that we are discussing is
only for the "hard cases" that do not work with simplier
training.

For example, the formal name of that female NASA astronaut,
"Heidemarie Stefanyshyn-Piper".

....called "Heide" by her friends, sometimes called "Piper"


Normal training yields "Heidemarie Stefanyshyn Piper".

Notice the missing hyphen just before "Piper".



Our convoluted training is necessary, to get the hyphen.


BTW, all speech apps have their virtues and drawbacks,
in an ideal world a speech app would combine all the good
points of the various present programs:

1) Vista Speech -
(formally "Windows Speech Recognition" or WSR)

2) Dragon (specifically, the "Pro" version 10.1)

3) "MacSpeech Dictate 1.3"
(bottom of the ladder, but only choice for Mac users)

Vista is by far the easiest to learn.

Dragon is the only choice for maximum productivity,
despite the loud protestations of Vista Speech advocates
to the contrary.

Dragon can be learned fairly rapidly, but a user has
to work harder at doing it.

The knowbrainer.com forum section is a good place
to pick up tips on how to master Dragon.



Speech recognition is a marginal human activity at best,
it is downright difficult to find good uses for speech apps.

Vista Speech just does not have all the productivity
enhancements of Dragon, a much more mature speech app.

This scarcity of features shoots down the professional use
of Vista Speech, which accounts for the reason why Dragon
still enjoys the overwhelming lions share of the
very lucrative speech recognition market.

Mark-


--
BTW, for Dragon users who are past the newbie stage,
there are 3rd-party video tutorials for sale, which
greatly ease the tough job of learning Dragon.

I typed into Google the search term:

Dragon NaturallySpeaking video tutorial



Unfortunately, Dragon users need this added help,
because of Dragon being harder to learn than
Vista Speech.

I have an old 3rd-party video tutorial I bought in 2006
for older version 9 Dragon, which I am presently reviewing
to refresh my understanding of allowing another person to
correct my dictation, thereby freeing me from this chore.

I enjoy unusual uses of Dragon, such as allowing myself
to dictate while I am jogging.

(using tiny Olympus DS-5000 recorder in concert with an
"ear-mounted" microphone)

Resulting accuracy is almost as good as dictating directly
to my computer.

On Sun, 03 May 2009 00:27:21 -0700, Mark Conrad wrote:
[color=blue]
> Well, I do not know what grade you teach in your job, so I will
> pretend I am you, and am teaching a bunch of graduate students
> about Einstein's problem understanding another genius of his
> era, a scientist named Paul Dirack, the man who predicted
> the existence of anti-matter.
>
> Okay, "Dirack" is a tough name for Dragon to handle, and I know
> that I will be constantly using that name while I am preparing
> my next-day lessons for my graduate students, using my
> handy-dandy Dragon speech app to prepare my lesson.
>
> I dictate "Paul Dirack" just for kicks, to see what Dragon
> will print. Dragon prints "pole direct". Bad Dragon !!![/color]

Bad Mark !!!

The man's name was Paul Dirac.

Usually given as P. A. M. Dirac.

So of course Dragon should have at least given "pole direc" as its first
guess :-)

--
Gene E. Bloch letters0x40blochg0x2Ecom

In article <l1d08lt74lp9$.nueccwm2mc5y$.dlg@40tude.net>, Gene E. Bloch
<not-me@other.invalid> wrote:
[color=blue][color=green]
> > Well, I do not know what grade you teach in your job, so I will
> > pretend I am you, and am teaching a bunch of graduate students
> > about Einstein's problem understanding another genius of his
> > era, a scientist named Paul Dirack, the man who predicted
> > the existence of anti-matter.
> >
> > Okay, "Dirack" is a tough name for Dragon to handle, and I know
> > that I will be constantly using that name while I am preparing
> > my next-day lessons for my graduate students, using my
> > handy-dandy Dragon speech app to prepare my lesson.
> >
> > I dictate "Paul Dirack" just for kicks, to see what Dragon
> > will print. Dragon prints "pole direct". Bad Dragon !!![/color]
>
> Bad Mark !!!
>
> The man's name was Paul Dirac.[/color]

You are correct, seems I had a senile attack!

Dirac had a heck of a time convincing his peers that there
might be such a thing as anti-matter, back in those days.

Mark-

On Mon, 04 May 2009 16:46:43 -0700, Mark Conrad wrote:
[color=blue]
> In article <l1d08lt74lp9$.nueccwm2mc5y$.dlg@40tude.net>, Gene E. Bloch
> <not-me@other.invalid> wrote:
>[color=green][color=darkred]
>>> Well, I do not know what grade you teach in your job, so I will
>>> pretend I am you, and am teaching a bunch of graduate students
>>> about Einstein's problem understanding another genius of his
>>> era, a scientist named Paul Dirack, the man who predicted
>>> the existence of anti-matter.
>>>
>>> Okay, "Dirack" is a tough name for Dragon to handle, and I know
>>> that I will be constantly using that name while I am preparing
>>> my next-day lessons for my graduate students, using my
>>> handy-dandy Dragon speech app to prepare my lesson.
>>>
>>> I dictate "Paul Dirack" just for kicks, to see what Dragon
>>> will print. Dragon prints "pole direct". Bad Dragon !!![/color]
>>
>> Bad Mark !!!
>>
>> The man's name was Paul Dirac.[/color]
>
> You are correct, seems I had a senile attack!
>
> Dirac had a heck of a time convincing his peers that there
> might be such a thing as anti-matter, back in those days.
>
> Mark-[/color]

Physics has gotten weird, this last 100 years or so :-)

I admit to problems with string theory and dark matter. I kind of like
neutrinos a lot, though :-)

--
Gene E. Bloch letters0x40blochg0x2Ecom