We invite you to participate in the shared task on Explainability of Cross-lingual Emotion Detection (EXALT), organized as part of WASSA 2024 at ACL 2024. The goal of this task is two-fold: developing highly performant models for cross-lingual emotion detection while also encouraging model explainability and interpretability.
**Task Description**
You can participate in three different tracks:
Track 1: Cross-Lingual Emotion Detection, which involves emotion classification with 5 labels with transfer from English to a multi-lingual test set that also includes French, Spanish, Dutch and Russian tweets.
Track 2: Trigger Word detection (binary), a task where participants have to identify which words in a text are most essential for expressing emotion. For this track, we treat trigger word detection as a binary classification task.
Track 3: Trigger Word detection (numerical), a task where participants have to identify which words in a text are most essential for expressing emotion. For this track, participants can provide numerical importance scores for each individual word.
For more information, please refer to the web page:
https://lt3.ugent.be/exalt/
The data is available through: https://codalab.lisn.upsaclay.fr/competitions/17730
**Important Dates**
15 February, 2024: first call for participation. Training and validation data released, CodaLab opens for team registrations.
13 May, 2024: Evaluation period/phase begins (= test data released)
15 May, 2024: Evaluation period/phase ends (= system submission deadline)
17 May, 2024: Announcement of final results
22 May, 2024: Submission of system description papers for review (we encourage participants to share models, code, fact sheets, extra data, etc. with the community through github or other repositories on paper publication.)
22 June, 2024: Notification of Acceptance
01 July, 2024: Camera-ready deadline
15-16 August, 2024: Presentations of the systems at the WASSA 2024 workshop (Bangkok, Thailand)
**Task Organizers**
Aaron Maladry
LT3, Ghent University
Pranaydeep Singh
LT3, Ghent University
Els Lefever
LT3, Ghent University
**Contact**
You can contact all organizers of the shared task at exalt-wassa2024 [at] googlegroups.com