Paper Format
The paper will be in the ArabicNLP (formerly WANLP) proceedings and must follow the conference template available at: https://arabicnlp2024.sigarab.org/call-for-papers
The paper is expected to be up to 4 pages long plus any number of pages for references. Please, note that the review process is not double-blind, so anonymity is not required.
The title of the system description papers should have the format <team_name at shared_task_name: title>, e.g., SuperHeros at StanceEval2024: All you need is Bla Bla Bla.
Authors' names should not be anonymous.
Submissions should be done via OpenReview: https://openreview.net/group?id=SIGARAB.org/ArabicNLP/2024/Shared_Task/StanceEval#tab-your-consoles .
Paper Structure
A common structure for system description papers includes the following:
Abstract: four/five sentences highlighting your approach and key results.
Introduction: ¾ a page expanding on the abstract mentioning key background such as why the task is challenging for current modeling techniques and why your approach is interesting/novel.
Data: review of the data you used to train your system. Be sure to mention the size of the training and test sets that you’ve used, and the label distributions, as well as any tools you used for preprocessing data.
System: This section provides a detailed description of how the systems were built and trained. If a neural network was employed, it should detail any pre-trained components, the model training process, hyperparameters chosen, and experimented variations. Additionally, include information on the duration of model training and the infrastructure utilized for training. While other paper styles may include background information as a separate section, it is acceptable to simply include citations to similar systems that inspired your work within the description of your system.
Note: Please consider releasing your code publicly to enhance reproducibility and facilitate faster learning for others. If you have already done so, kindly mention it in your paper and provide a public link.
Results: a description of the key findings of the paper. If additional error analysis has been conducted to identify the types of errors the system makes, this information is invaluable for the reader. Unofficial results obtained after the submission deadline can also be highly beneficial.
Discussion: a general discussion of the task and your system. It should include a description of characteristic errors and their frequency over a sample of development data. Additionally, consider discussing what you would do if you had another three months to work on it.
Conclusion: a restatement of the introduction, emphasizing the insights gained about the task and how to model it effectively.
Please use the following BibTex entries to cite the shared task paper and the dataset paper:
@inproceedings{StanceEval2024,
title={StanceEval 2024: The First Arabic Stance Detection Shared Task},
author={Alturayeif, Nora and Luqman, Hamzah and Alyafeai, Zaid and Yamani, Asma},
booktitle={Proceedings of The Second Arabic Natural Language Processing Conference (ArabicNLP 2024)},
year={2024}
}
@inproceedings{alturayeif2022mawqif,
title={MAWQIF: A Multi-label Arabic Dataset for Target-specific Stance Detection},
author={Alturayeif, Nora Saleh and Luqman, Hamzah Abdullah and Ahmed, Moataz Aly Kamaleldin},
booktitle={Proceedings of the The Seventh Arabic Natural Language Processing Workshop (WANLP)},
pages={174--184},
year={2022}
}