SightCAT: a market survey for dictation software targeted at professional translators
Thread poster: John Moran
John Moran
John Moran  Identity Verified
Ireland
Local time: 11:34
German to English
+ ...
May 8, 2018

Hi all

My former language technology research group in the ADAPT Centre in Trinity College Dublin is currently carrying out a market research survey on the use of speech recognition and human transcription by professional translators here:

http://www.sightcat.net

If you use Dragon Naturally Speaking or any other speech recognition software with a CAT tool or word processor i
... See more
Hi all

My former language technology research group in the ADAPT Centre in Trinity College Dublin is currently carrying out a market research survey on the use of speech recognition and human transcription by professional translators here:

http://www.sightcat.net

If you use Dragon Naturally Speaking or any other speech recognition software with a CAT tool or word processor it might be interesting as the idea is to provide a more accurate speech recognition system that has a range of features targeted at translators. These would include better tag support and better accuracy.

The idea to is to provide it as a service inside existing CAT tools (a bit like MT) but before that can happen they need to build a picture of whether there is any demand.

If you translate into a language that Dragon does not support and other SR options are unsatisfactory (as they don't adapt to the text or the speaker) it is well worth taking part as the order in which new languages are developed will probably depend on the number of responses they get on the survey for that target language.

There is a short video to explain the concept and a short survey that takes about 5 mins on the site.

It closes soon so please do take 10 mins to look it if you are even vaguely interested in speech recognition.

Kind regards,

John Moran

John Moran
Transpiral Ltd.
Collapse


 
Dan Lucas
Dan Lucas  Identity Verified
United Kingdom
Local time: 11:34
Member (2014)
Japanese to English
Interesting May 8, 2018

John Moran wrote:
It closes soon so please do take 10 mins to look it if you are even vaguely interested in speech recognition.

Thank you, done.

Dan


 
Mirko Mainardi
Mirko Mainardi  Identity Verified
Italy
Local time: 12:34
Member
English to Italian
Done May 8, 2018

Not sure how they would handle the tag issue, but it'd be nice if they found a viable solution.

Also, it seems they're also interested in pricing, although I'm not sure a per word model is the best way to go (as opposed to a license, timed or perpetual). But that's just me, of course...


 
John Moran
John Moran  Identity Verified
Ireland
Local time: 11:34
German to English
+ ...
TOPIC STARTER
Pricing May 8, 2018

As far as pricing is concerned that is one of the things the survey is looking for information on.

I'm advising on the market research on a pro bono basis so I'm not speaking on behalf of the team. With that caveat in mind, my feeling is that per word (i.e. usage) pricing is preferable to any other model for ASR and it should be based on the utility of the system relative to competitors (if there are any ASR systems for that language). Pricing should also take into account the typi
... See more
As far as pricing is concerned that is one of the things the survey is looking for information on.

I'm advising on the market research on a pro bono basis so I'm not speaking on behalf of the team. With that caveat in mind, my feeling is that per word (i.e. usage) pricing is preferable to any other model for ASR and it should be based on the utility of the system relative to competitors (if there are any ASR systems for that language). Pricing should also take into account the typical word prices charged for a target language, e.g. Swedish would be more expensive than Polish.

Obviously, for manual transcription pricing depends more on the cost of production and it would only work economically for translators who can up their productivity and sales volume enough to pay for the 2nd person in the loop. Anecdotally though, there are translators out there who do that already and the benefits are sometimes dramatic.





[Edited at 2018-05-08 12:10 GMT]
Collapse


 
John Moran
John Moran  Identity Verified
Ireland
Local time: 11:34
German to English
+ ...
TOPIC STARTER
Tags May 8, 2018

Mirko Mainardi wrote:

Not sure how they would handle the tag issue, but it'd be nice if they found a viable solution.

Also, it seems they're also interested in pricing, although I'm not sure a per word model is the best way to go (as opposed to a license, timed or perpetual). But that's just me, of course...


I understand tags would be statistically placed using word alignment data. For the rare cases where the tag is misplaced the translator can just fix the placement. There would also be a command to place the next tag for people who prefer more manual control. It would not work well for tag forest or deeply embedded tag situations but that is an unusual case and normally a sign of poor upstream file engineering (e.g. ocr'ing pdfs).


 


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

SightCAT: a market survey for dictation software targeted at professional translators






Wordfast Pro
Translation Memory Software for Any Platform

Exclusive discount for ProZ.com users! Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value

Buy now! »
TM-Town
Manage your TMs and Terms ... and boost your translation business

Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.

More info »