@
We are excited to announce the
Arabic Sentence Segmentation Shared Task, which focuses on automatically identifying sentence boundaries in Arabic documents. The task is formulated as a binary token classification problem: given an Arabic document as input, systems must predict whether
a sentence boundary follows each token.
Task 1: Paragraph-Aware Arabic Sentence Segmentation
Given an Arabic document with its paragraph boundaries, predict for each token whether a sentence boundary follows it.
Task 2: No-Punctuation Paragraph-Aware Arabic Sentence Segmentation
Given an Arabic document with its paragraph boundaries but without punctuation, predict for each token whether a sentence boundary follows it.
Task 3: No-Paragraph Arabic Sentence Segmentation
Given an Arabic document without its paragraph boundaries, predict for each token whether a sentence boundary follows it.
Task 4: No-Punctuation No-Paragraph Arabic Sentence Segmentation
Given an Arabic document without its paragraph boundaries and without punctuation, predict for each token whether a sentence boundary follows it.
For each task, there will be two tracks, allowing different data sources for training:
Closed and Open. Participants may compete in any combination of subtasks and tracks.
Important Dates:
All deadlines are 11:59pm UTC-12 (anywhere on Earth):
-
June 1, 2026: Release of training, dev and open test data, and evaluation scripts.
-
July 20, 2026: Registration deadline and release of test data.
-
July 25, 2026: End of evaluation cycle (test set submission closes).
-
August 8, 2026: System description paper submissions due.
-
August 15, 2026: Notification of acceptance.
-
August 15, 2026: Final results released.
-
August 22, 2026: Camera-ready versions due.
Awards:
Top-performing Systems:
Best System Description Paper:
-
We will also award a
$200 prize for the best system description paper, recognizing clarity, technical quality, reproducibility, and insight, independent of shared task performance.
Organizers:
-
-
-
-
Bashar Alhafni: Mohamed bin Zayed
University of Artificial Intelligence
Contact:
For any questions related to the task, check out the
FAQs. Feel free to post your questions on our
Slack workspace. You are also welcome to contact the organizers directly at this email address:
araseg26....@aramlab.ai