SemEval 2025 Task 10: Multilingual Characterization and Extraction of Narratives from Online News

Event Notification Type: 
Call for Participation
Abbreviated Title: 
SemEval 2025 Task 10: Leaderboard Open

Dear Colleagues,

We are excited to announce that the Leaderboard for SemEval 2025 Task 10: Multilingual Characterization and Extraction of Narratives from Online News is officially open!

Task Overview
The task focuses on analyzing multilingual news articles in two critical domains—Ukraine-Russia War and Climate Change—and is subdivided into three subtasks:

1. Entity Framing
- Goal: Assign fine-grained roles (e.g., protagonist, antagonist, or innocent) to mentions of named entities in a news article.
- Task Type: Multi-label, multi-class text-span classification.

2. Narrative Classification
- Goal: Assign appropriate sub-narrative labels to a news article using a two-level taxonomy of narratives.
- Task Type: Multi-label, multi-class document classification.

3. Narrative Extraction
- Goal: Generate a free-text explanation (up to 80 words) for the dominant narrative in an article, grounded in evidence from the text.
-Task Type: Text-to-text generation.

This task provides an opportunity to push the boundaries of multilingual NLP and tackle challenges related to narrative understanding and extraction.

Important Dates
- Development Set Available: November 11, 2024
- Leaderboard Open: November 11, 2024
- Test Set Release: January, 2025
- Submission Deadline: January, 2025 (keep monitoring the website for precise dates)

How to Participate
1. Visit the official task website: https://propaganda.math.unipd.it/semeval2025task10/
2. Register your team to get access to the data and begin evaluating your models.
3. Submit your predictions to the leaderboard to compare with other participants.

This is a great opportunity to contribute to cutting-edge research in multilingual narrative analysis and to engage with the SemEval community.

We look forward to your participation!
Best regards,
The SemEval 2025 Task 10 Organizers