Who can see your viewing activity?
will this be recorded?
how can i get the recording after please?
Hi Crhis, we will make it available here: https://gourmet-project.eu/project-output/user-event-2021/
there’s an option to share screen with audio
The translation seemed fast enough, I would say
Thank you, both of you
I have to leave. Sorry. Thanks for comming.
Thank you Felipe, bye bye
Mo McRoberts @ BBC Datalab
ich muss los! but fascinating work & really interesting presentations — i'll definitely be looking at some of the downloadable assets later on — danke sehr! 🙂
I have to go. Have a nice day!
Bye bye to those who leave! Thanks for joining us today.
For those of you in BBC, you can try out LPT here https://live-pages-monitor.tools.bbc.co.uk/
Juan Antonio Pérez
It was not only me :-)
Frank uses Serbian, Tamil and Tigrinya models from GoURMET.
Again for BBC colleagues (apologies!) you can try out Frank here https://newslabs-frank.tools.bbc.co.uk/
And Multilingual graphical storytelling herehttps://gst.tools.bbc.co.uk/
Thanks everyone - great demos and great to see so much good work happening to support multilingual journalism and content creation - have to go now!
Thanks all for the session .. I need to drop off now. All the best 👍🏻
What does multilingual journalism mean for you?How can we work better in a truly multilingual environment?Where would multilingualism generate most value for audiences?
Great demonstrations that really make visible the potential benefits. I was wondering if DW or BBC have empirical quantitative data about the benefits. For example, time saved translating from scratch compared to starting from machine translations?
In your BBC GST tool, a) do you use machine learning to choose the right graphics?
b) do you improve the model on the fly based on people choices of the alternatives?
@Eyal Serbian team have reported it cut their reversioning time to a third (not by a third)But we need more quantitative data
c) how many graphics do you have in your library? And is the library regularly extended?
These are the systems we used for entity extraction in GST: SpaCY, TF-IDF (using TextBlob) and Amazon Comprehend.
Please contact us if you have more questions, comments or want to use our open-source models: firstname.lastname@example.org
Regarding Live Pages: Did you find a way to measure how the tool helped improve the discovery of stories that would have been otherwise missed by interested language services?
Apologies my mic has failed
And so has my speaker
Fascinating to see all the possibilities for production teams. Many thanks. Estelle, BBC French language
Thanks for the update/info. Sorry need to go to another Zoom meeting.