Official account of the Touché / @SemEvalWorkshop 2023 Shared Task 4. ValueEval: Identification of Human Values behind Arguments.
Joined July 2022
- Tweets14
- Following42
- Followers28
- Likes13
Dataset update! zenodo.org/doi/10.5281/zenod…
67,352 sentences in 9 languages, annotated for 19 values. First access to data from ValuesML (teapawt.net/EU_ScienceHub/status/1… @EU_ScienceHub)!
And don't forget to register to our shared task: touche.webis.de/clef24/touch… @CLEF24Grenoble #DataForGood
Understanding citizens' values through #MachineLearning
70+ researchers are joining forces to bridge the gap between citizens' values and political aspirations. Learn more about the project: knowledge4policy.ec.europa.e…
#Science4Policy
ValueEval retweeted
ValueEval'24 is part of Touché, which just opened registration!
This time we have long texts, multi-lingual, and general awesomeness thanks to collaborating with the ValuesML campaign: teapawt.net/EU_ScienceHub/status/1…
Stay up-to-date via our mailing list: groups.google.com/group/valu…
Working in Argumentation? Time to participate in Touché 2024!
Three shared tasks:
- Human Value Detection
- Ideology and Power Identification in Parliamentary Debates
- Image Retrieval/Generation for Arguments
Submission deadline is May 6th!
More info: touche.webis.de/
Thanks to everyone for attending the talk yesterday at @SemEvalWorkshop! If you have not yet, go to valueeval.webis.de and sign up for our mailing list to stay up-to-date for ValueEval'24!
In case you are at @aclmeeting, here are your time slots for learning more about the ValueEval participants at @SemEvalWorkshop (in two days!). Hope to see you there!
Final dataset is ready and early bird deadline in 10 days.
Time to get your model flying!
valueeval.webis.de
ValueEval retweeted
Registration is now open for the 4th shared tasks on Argument and Causal Retrieval, Touché @ CLEF'23!
- 4 Tasks (until May 2nd 2023)
- Help us expand our image dataset (until Dec. 31 2022)
Learn more: touche.webis.de/
Register: clef2023-labs-registration.d…
@clef_initiative
Very interesting thread. What are the human values resorted to on both sides of the discussion? (clearly one category is "achievement")
Do you think we should add a few hundred of the tweets to our dataset?
(small reminder to our call for data: touche.webis.de/semeval23/to… )
We are still open to integrate more data sources into our dataset for diversification.
Put differently: we offer free annotation of text data (premise-claim pairs) acc. to the value taxonomy.
Check touche.webis.de/semeval23/to… and contact us! Thanks to all who already did!
Don't know where to start? Watch our new 4-minute introduction! youtu.be/ZAQ4LELCCY4
@SemEvalWorkshop #semeval #touche
What does it look like if a team takes on our @SemEvalWorkshop human value detection task? We asked @OpenAI's DALL-E for an abstract painting just so you know what you can expect from participating:
ValueEval retweeted
Task 4: ValueEval: Identification of Human Values behind Arguments
valueeval.webis.de
@ValueEval
Task is up and ready: valueeval.webis.de
Join us for @SemEvalWorkshop 2023 and learn more about human values in argumentation on the way.
We already have 5K arguments annotated.
Registration is open (forms.gle/ib72JUtsxhb3ZMNw5). So: