ROUGH EDITED COPY

NTID/RIT

SPEECH-TO-TEXT SERVICES NETWORK: A NEW PROFESSIONAL ORGANIZATION

DENISE KAVIN

JUNE 27, 2005

10:00 A.M. CST

CAPTIONING PROVIDED BY: CAPTION FIRST, INC.

PO BOX 1924

LOMBARD, IL 60148

* * * * *

This is being provided in a rough-draft format. Communication Access Realtime Translation (CART) is provided in order to facilitate communication accessibility and may not be a totally verbatim record of the proceedings

* * * *

>> Good morning!

I know the presentations have been delayed because of the presentation that started in Panara.

So we have a wonderful speaker this morning, Denise Kavin.

I want to ask if you would please fill out the evaluation form at the end of the presentation, and then give it back to me.

Thank you.

>> DENISE KAVIN: I'm sorry for the late start, but I think that I can condense my material and my presentation a little bit.

People will be a arriving like they are right now, but that's all right.

We'll give them another minute.

>> DENISE KAVIN: I work for PEN-International.

That's the organization hosting this tech symposium.

I am associated with PEN-International.

I just started six months ago.

Before I came to PEN-International I worked as coordinator of deaf and hard of hears at community college in the Chicago area.

At that program I was very active with speech-to-text services.

I helped set up the captioning program there for our deaf and hard-of-hearing students.

At that time I was asked to become the chairperson of STSN, and I was accepted, and I accepted that position, and I brought that responsibility with me here to Rochester.

Okay?

Feel free to ask questions during the presentation.

If you don't understand what's going on, please feel free to put your hand up.

We have interpreters for both voice and sign.

You don't have to take any notes.

I have handouts with everything on it for you.

>> DENISE KAVIN: I want to ask you guys, how many of you work at colleges or are students?

You are a student?

You work?

You work at a college?

Work at college.

>> Syracuse University.

>> DENISE KAVIN: It will definitely apply to you.

Are there others?

What kind of --

>> I work at the U.S. Government printing office.

>> DENISE KAVIN: I am going to wait another minute here.

People are wandering in.

I was just introducing myself, that I came from a community college, and I work for PEN-International now.

And feel free to ask questions throughout the program.

I'm here to talk about STSN, speech-to-text services network.

It's a new organization that was just set up, maybe about a year ago, it was set up, and I am the first president, or chairperson, of this organization STSN.

So what is STSN?

STSN is for people who provide captioning services in all different environments, CART, like a stenographer in court, C-Print-Pro, TypeWell, automatic voice recognition technology, all of those various technologies.

I will talk about the differences between them in a minute.

RID requirements for interpreters, certification for interpreters of the deaf.

But for captionists there is no such organization set up, so that's why we set up the STSN because captioning services is really, really growing, and I'm sure you've noticed that the growth is increasing quickly.

And we need to provide some support for captionists.

There's a little bit of training involved.

I was asked if the training was included.

We just had our first official training in May in Chicago.

We're going to have another training session next year in the spring.

>> DENISE KAVIN: Why do we need a professional organization?

The bottom line is there's no support for captionists anywhere out there.

For CART, those providers have a professional organization.

NCRA, the national court reporter's association, but there is nothing out there for C-print organization for TypeWell providers.

We need access to information.

We need resources.

We need to know what's the best practice out there and what are the guidelines.

How many of you subscribe to PEPNet newsletter on e-mail?

A few of you.

Everyday you get a brief summary of questions and discussions, concerns, and answers that are on list serve.

We notice that more and more questions keep coming up about captioning, and how do we find people to provide the captioning in the classroom for students?

And how can we get people to stay and not leave the job?

I mean, how can we find trained people?

What do we pay them?

What is the rate?

How are they qualified?

Do they fit with the student?

There are so many questions that come up related to this issue, and these questions keep coming up over and over again because there's no standard.

There's no training, no uniform training.

There's also a hot discussion now about the purpose of STSN.

Captioning services is growing, and it's growing fast.

So should we use captioning services as a way to provide services when there is a lack of interpreters, or is it a best match for the student?

I remember that in my community college where I was working before we didn't have enough interpreters, and, of course, we never have enough interpreters.

And then we found people who are trained as captionists, and we brought them in, but some people didn't feel it was effective for them because it wasn't a good match for the students.

Captionists can be good match for a student who maybe are not skilled signers.

Or who have a strong English skill and prefer to read the text and read the information instead of going through a Sign Language interpreter.

It also might be good for a highly technical course where there's technical vocabulary.

In our experiences with the captionists, they're not always a good match.

Sometimes like in a math class, for example, how do you caption math?

Or even for English classes, I mean, that's kind of surprising that they're not always a good match in the English classroom.

I was asked why.

I'm not exactly sure why.

I think that maybe because of the English is exact with the grammar and syntax, the noun, verbs, pronouns, adjectives and adverbs.

There are many interpreters and captionists that don't feel it's a effective way to convey the information.

Interpreter s are more effective way of conveying that information than English classroom.

Let's talk about what is speech-to-text services.

Speech-to-text is a format of realtime access, meaning that the student, the person gets access to the information at that moment.

That information from the presenter, like a note taker, it's not exactly the same.

With note taking, they're just summarizing and then the student reads it later.

It's not at that time.

It's not realtime.

Interpreter is realtime.

You have immediate access to the information there.

Speech-to-text is also immediate.

You have immediate access to the information at that time.

Many people often misunderstand the speech-to-text.

They will equate speech-to-text with one specific service like C-print, or TypeWell.

But really C-print and TypeWell are just different technologies, but they all have the same result.

They're all under the speech-to-text umbrella.

Sometimes people say, "Oh, C-print, C-print,"Żbut really they're talking about speech-to-text.

Speech-to-text is a general category, and then underneath that category there are these different systems.

>> DENISE KAVIN: This is what I was just talking about.

This presentation is not really about the technology.

I'm not knowledgeable about the technology, but I'm going to talk more about the organization, the speech-to-text.

It's a newly-developed organization.

That's why I am going to explain why we need an organization and why we need to have that in place.

I was asked to be the chairperson of this organization because of my organizational development skills, because of my neutrality.

I'm not affiliated with C-print, and I'm not affiliated with TypeWell, and I'm not affiliated with CART either.

I have very little experience with each of these different systems.

>> DENISE KAVIN: STSN, again, is realtime.

It's realtime access, but it can also be used as a note-taking services for later on.

It's a very unique way of providing note-taking services.

Are the interpreters following okay, the Russian interpreters?

Do you want me to slow down?

>> Yes, can you please slow down just a bit?

And may I ask you a question?

>> DENISE KAVIN: Yes, go ahead.

>> I have a question.

Speech-to-text, what you do mean?

Speech like verbal speech to text, or speech like Sign Language speech to text?

>> DENISE KAVIN: Anything that is verbal.

Anything spoken to caption.

>> Thank you.

>> DENISE KAVIN: It could be Sign Language to printed word and text as well, like what we have on the screen over here.

This is a perfect example.

I'm going to slow down a little bit because we have a Russian interpreter here, so I have to pace myself.

>> DENISE KAVIN: There are a few different systems.

One is CART.

Computer assisted realtime transcription.

CART provides intensive training.

It take as full two years, and it's a very intensive program.

They work in court settings, and their work is very high level.

It's verbatim.

It's exact word for word including words such as "um,"Żand any exact word that's included.

C-print is a system developed here at NTID.

It's based on phonetics.

The software is based on phonetics, and the user goes through maybe a one-week training period, and it's all online.

It's based on phonetics.

Would you mind?

Come on up.

She is directly involved with the C-print.

>> Audience member: I work with northeastern technology assistance, C-print organization is part of NSC and NTID, and I've seen many new people who want to go for the training for C-print.

It's a 10-week training.

So that's it.

>> DENISE KAVIN: Thank you.

So that was about C-print.

I like to call it a cross between CART and note taking.

And it's not a verbatim process.

It's pretty close.

The more skilled you become at C-print, the more the message becomes verbatim.

There's also another system called TypeWell.

That was invented by a person, a woman, from California, and she also started C-print.

Then she left and went back home to California, and she developed her own software.

And the end results are really the same.

It's just that the software itself is a little bit different.

It's based more on abbreviations.

It's about three days of face-to-face personal training with about 40 hours of preparation time that must be done before the training, one-on-one training.

So people often confuse C-print with TypeWell, but I say that the end result is really the same.

It's just different software involved, and different training opportunities.

TypeWell tends to be more popular out west, more in that part of the country.

In C-print, that tends to be more popular in the east because it was developed here at NTID.

And where I worked before at the community college, we had both, C-print and TypeWell.

But then it became much too complicated because we had toe have laptop -- had to have laptops installed in the classes for TypeWell with that software, and then different laptops for C-print with that software.

So it became much more complicated, and we made the decision to just pick one, and we went with TypeWell.

But both were equally good.

Okay.

There are a few possible different setups.

For students and for the captionist, they can sit side by side, and they can read the screen at the same time.

Or another setup if there were a few deaf students with a large television monitor that could be attached to the captionist's computer so that they could read the larger screen all at the same time.

Or they could use a wireless setup with two different computers in the same room.

Or it could be set up as a remote setup.

Remote setup means that the captionist is in an entirely different state.

For example, we have one class that happens in Illinois and gets -- the captionist is in the next state, in Wisconsin.

And some workshops here today and tomorrow will be captioned from Colorado, not by captionists here like we have in this room, but other workshops have a remote captionist.

It's really a wonderfully-amazing process.

Sometimes the low calorie sources are very scarce and there's not enough training.

So we contract out to other captionists in other states, and we have the technology that can support that.

Especially for schools.

Does anyone have questions?

Yes, over here?

>> Audience member: I have two questions for you.

They're quite simple questions.

The individual over here that's typing --

>> DENISE KAVIN: You are doing CART, right?

Yes, that captionist is doing CART, which is verbatim.

>> Audience member: Before you mentioned the training that C-print training, was that one week?

>> DENISE KAVIN: It's online it used to be one week, but now it's changed to be an online training which you do at your own time and at your own pace.

It doesn't necessarily have to last a week.

It could be done more quickly or more slowly.

Is that right?

>> Audience member: I'm not actually involved with C-print, but they can provide you with the software, and you can train at home, you know, during the night, or you train until you become comfortable with C-print, and then you can go to a 10-week training program and go through that intensive program.

But before that, you need to get the basic knowledge of the technology on your own.

>> Audience member: Oh, okay, so do you mean a 10-week training program that you are in a classroom setting?

>> Audience member: Yes, that's what I mean.

>> Audience member: One on one?

>> Audience member: Correct.

I'm talking about in another state.

But here at NTID we're in a group, and it's 10 people, 10 to 15 people, that get together and they go through the course together.

>> Audience member: Oh, thank you very much.

>> DENISE KAVIN: But for people from other states they also can be trained online, not in person, right?

Oh, did you have a question?

>> Audience member: I was trained the old --

>> DENISE KAVIN: Hold on.

She would like to say something.

>> Audience member: I was trained the old-fashioned way before the online was done, whereas we would go wherever the training was for a week and then we had five weeks of our own intensive training that we had to report back to our coach or team leader so it is intensive.

Now they've changed it, and it's one week online, but there is still the captionists that have to really practice before they can actually begin to caption.

They have about five or six weeks that they have to really, really practice intensively before they get it.

>> DENISE KAVIN: Oh, thank you.

Did you have another question you would like to ask?

>> Audience member: About the TypeWell training.

>> DENISE KAVIN: That's C-print, right?

I have a question back here asking about TypeWell.

>From my experience there's a 40-hour preparation period that happens before the training, and that's online.

Then there is a three-day training in person that happens all day, and it would happen in whatever part of the country that person was.

Now the people providing the training for TypeWell -- well, one of the drawbacks is that there is only one person providing the training, whereas C-print has many different people who provide the training.

TypeWell has just one or a few that you actually have to travel to go and get the training.

>> Audience member: Thank you.

>> DENISE KAVIN: Do you have a question?

>> Audience member: One question right now.

All of the systems that you are talking about are based on a keyboard entry.

>> DENISE KAVIN: Yes.

>> Audience member: There are some systems that use speech recognition software.

Are you working with those also?

>> DENISE KAVIN: Yes, thank you.

I was going to mention that.

Automatic speech recognition systems, that's also included in this, but from my experience so far we have a lot of people -- we don't have a lot of people as involved with automatic speech recognition.

That hasn't happened yet.

We're seeing less of that now.

So I'm not seeing a lot of automatic speech recognition systems, but I think that it will grow, and I predict that it will expand and more and more people will become involved.

>> Audience member: We tried an automatic speech recognition software with a course.

The biggest downfall was the training of the professor with the software.

Technology worked great, but training by the instructor fell short.

>> DENISE KAVIN: Was it resistance on the instructor's part?

Was it just a lack of information you do think?

>> Audience member: It was a detailed course which the instructor did not take the time to train the software to recognize his voice to pick up the new vocabulary, the new words that were coming in.

>> DENISE KAVIN: Oh, okay.

Generally what I've seen, there are two ways: One is a speech recognition with a shadower, and that would be a person who sits and repeats everything that's been said, and they wear a special sort of mask.

It look as little bit odd it looks like a gas mask, to tell you the truth.

But there is also another kind where the professor speaks directly for themselves and they wear a microphone.

So I'm not sure exactly how accurate that information would be.

A question?

>> Audience member: In regard to his comment that requires the professor to train, does that really fall into this category of organizing, you know, typists and such?

Because that's an automated system that won't need a typist.

It's just strictly the professor and software.

>> DENISE KAVIN: They don't need a typist.

Or they could use a shadower.

>> Audience member: Right.

The problem that we've experienced on our campus using the shadower, and even using the mask is that students that sit around the person using the mask complain about whispering, or muffled sounds.

>> DENISE KAVIN: Oh, thank you.

In fact, I had some personal experience with it at home.

We have a special -- it's like a captioned phone.

It's called CapTEL.

Any of you heard of that?

My husband is deaf, but he speaks very well.

When he makes phone calls he likes to speak for himself.

So he picks up the phone, and he dials the number, and then becomes hooked up to an operator.

The operator will repeat everything that he said, and then the phone would actually caption what the other person is saying.

But there are so many errors involved, and the person that he is talking to can actually hear the whisper of the shadower repeating everything that he is saying.

So there is still a lot of things that need to be worked out with voice recognition.

But we have some demonstrations downstairs right now by caption mic, and it's a speech recognition technology, and it's right there.

So afterwards feel pre to go down and talk with them.

>> DENISE KAVIN: Okay.

I'm not saying that I have all of the answers, but these are the questions that have repeatedly come up.

And these are the reasons that we set up speech-to-text service network.

And the questions actually we're trying to make services become more uniform because people often ask, "How you do match the student with the captionist?"

"how you do prove that can benefit from captioning?"

One time I had a student that was a wonderful signer, and they wanted captioning.

And I had to turn down that request because I couldn't justify the cost versus the need.

So that student did very well with an interpreter, so you have to figure out how to justify that.

And where do people get their training?

Right now, people are just going wherever the train something happening, and they travel all over the country for that they watch the list serve for the information to come out with the announcements, and then they just grab the training.

There's no coordinated effort right now.

And there's a lot of confusion, too, between CART services, C-print services, and TypeWell services, and which one the student should pick.

And how can you know if the captionist is actually qualified?

There is no way to evaluate them.

There's no certification right now.

So we just have to get by and hope.

And there's one huge thing that I've discovered is that there are people willing to be trained.

In the past, some people had luck with that, and they used more clerical staff instead of interpreting staff, or interpreting students were used, but it wasn't a great match.

And then we had to figure out pay.

We pay them less than interpreters because that's very awkward?

We don't have an answer on what the rate should be.

Right now they generally pay between $20and $30per hour.

But that's very generally.

There is no uniformity across the country.

Do you have a question?

>> Audience member: I have a question about the TypeWell student's needs, C-print, and the various others.

Do they group ASL users together?

Students who use both modes of communication?

Cave there are two or three different groups, students who don't know Sign Language, that would be the priority group.

Students who are just not skilled in English.

They're in a group -- they can't read fast because their English is not strong enough to be able to read the text and follow along.

So they're not strong candidates.

Or perhaps classes that are mainstream with a large number of deaf students in them they use interpreters plus captionists.

We like that a lot!

There are three, four, five different deaf students in a class, and they go over the notes afterwards, and there is a tremendous benefit to them.

And students with other learning disabilities also access these services.

And the courses that are really heavily lecture, history courses, literature courses, they tend to be really good classes because they're so lecture-based.

They're good classes to put in captionists as opposed to like a lab or a math class or an English class.

They're less of a good match for a captionist.

Do you have a question?

>> Audience member: Why would you prefer that --

>> Interpreter: Hold on a minute.

>> DENISE KAVIN: He just asked why it's not good for a math or engineering class, why there is not a good match?

That surprised you?

For those types of classes, it's a better match for the history.

For math, because they use numbers, how you do caption the numbers?

You know, the formulas, complicated equations, there's not a good way to put that down.

You know, with numbers, and equal signs, and greater and less thans, and engineering as well because of the terminology.

It's not been very effective.

But thanks for your question.

Is everyone doing okay?

We've got about five minutes remaining.

We do have a new website.

It was recently updated just last week.

I will give it to you at the end.

It's also going to be on your handout.

It's a very good website.

There are a lot of questions and answers, and technical information included there.

>> Audience member: I recently got on the Internet and I was looking at the website, and I wanted to join but there was a problem.

Have you been experiencing difficulties?

>> DENISE KAVIN: Yes.

>> Audience member: Okay.

So it wasn't me.

(Laughter)

>> DENISE KAVIN: No, no.

Right now, we don't really have a membership because we just applied for accreditation, and I just mailed out the paperwork to get that, and we haven't been formally accepted.

We're not exactly sure what to do about membership dues or sponsors, supporters.

We will accept donations, but currently we're looking at the membership structure and trying to decide what kind of membership categories we should set up.

So in the future we will be accepting supporters, and we have over 100 supporters to date.

>> DENISE KAVIN: STSN, the concept was developed in 2002 at a conference in Kansas City.

And then at a conference just last year in Pittsburgh, Pennsylvania, it was established as formal organization.

And these are a list of names of the people who were involved in that effort in setting up this organization from the very beginning.

And it's a varied group.

People come from all various backgrounds.

This includes people from C-print, automated speech recognition, and TypeWell.

High educational students for PEN-International for personal services for education and PEPNet.

>> DENISE KAVIN: And this is a list of our goals.

To provide information, and to remain neutral.

Not to support one system over another.

We try not to do that.

To try and develop cutting-edge -- oh, to provide leadership for providing certification and education and professional development because currently we don't have these opportunities.

We would like to establish them into a more uniform way.

, Do you have a question?

>> Audience member: In regards to the various types of software, you know, obviously C-print, TypeWell, and such.

They're similar in how they operate but they're very different in methodology.

How are you going to be able to credential the various nature of each of these categories?

And make a standard one?

>> DENISE KAVIN: We struggle with that internally.

You're right.

I'm not sure yet.

C-print and TypeWell, and even CART, we're struggling with that issue right now.

So, I'm sorry, but I don't have a more specific answer for you.

>> DENISE KAVIN: Right now we have an advisory board and a committee, and they have lists.

So if you are interested in joining a committee and provide some assistance, we would happily accept that.

Just see me after this presentation and give me your name and e-mail address, and I would be very happy to include you in one of our committees.

They're pretty active right now, too.

And we're all over the country, so right communication is tending to occur through e-mail.

We don't have the funding yet to have more face-to-face meetings, but we're hoping to have larger national conventions, and then at the conferences we would get together, have our face-to-face meetings, and have our discussions.

I'm going to have to start to wrap up because I need to give you a little break before you move on to your 11:00 presentations.

This is the advisory board, the list of names.

And the people from here would be -- oh, we have people from here, from California, from Connecticut, from Arkansas, and this group was planning to get together in spring 2000, and that's how long our term lasts, until the spring of 2006.

This is a list of our current committees.

If you are interested volunteering to be on a committee, just let me know.

I would be very happy to take your name and get your e-mail address.

And tap into your knowledge and get your feedback.

Are there any remaining questions before we wrap up?

Okay.

Thank you very much!

(Applause)

Enjoy the next presentation.

Thank you again.

>> Thank you very much, Denise.

I am going To remind you to please fill out the evaluation forms, this yellow sheet.

And then give them back to me, okay?

>> Audience member: During this presentation, did you have any program this working right now?

I mean, special computer program.

Was there any special program working right now from speech-to-text right now during this presentation?

Did we see that in our --

>> DENISE KAVIN: Yes, over here on the screen.

>> This is CART.

>> DENISE KAVIN: What's CART is that you what are asking see? Do you want to talk after the session one on one?

>> Audience: Individually, can we ask you?

>> DENISE KAVIN: Yes, that will be fine.

I will stay.

I will be here, and we can talk right after this.

>> Audience member: Thank you.

* * * * *

This is being provided in a rough-draft format. Communication Access Realtime Translation (CART) is provided in order to facilitate communication accessibility and may not be a totally verbatim record of the proceedings

* * * *