Call for Participants: The 1st AraFinNLP Shared Task 2024

Shared-task URL:  https://sina.birzeit.edu/arbanking77/arafinnlp/
Registration: https://forms.gle/rAZkRuiiEdjzTDedA


We are excited to announce the first Arabic Financial NLP (AraFinNLP) shared 
task, which aims to advance Arabic Financial Natural Language Processing (NLP). 
In today's dynamic financial landscape, the Arab world is witnessing 
significant growth in its stock markets and financial sectors, highlighting the 
pivotal role of Financial NLP in this context. AraFinNLP shared task is part of 
the 2nd ArabicNLP conference, which is co-located with ACL 2024 in Bangkok, 
Thailand, scheduled for August 16, 2024.

Shared Task Description

The AraFinNLP shared task comprises two subtasks:

Subtask 1: Multi-dialect Intent Detection

Subtask-1 involves Cross-dialect Intent Detection in the banking domain. This 
subtask is centred around developing models that can accurately classify 
customer intents from queries in various Arabic dialects. Participants will be 
provided with a dataset containing queries in English and their corresponding 
MSA and Palestinian translations. The challenge lies in training models that 
not only understand MSA and Palestinian but are also adept at interpreting an 
array of Arabic dialects, such as Gulf, Levantine, and North African. This 
subtask aims to foster the creation of NLP tools that can seamlessly interact 
across the diverse linguistic landscape of the Arab world, enhancing customer 
experience and operational efficiency in the banking sector. The evaluation 
process for this subtask, combined with the second subtask of Dialectical 
Translation for Intent Detection, will be designed to comprehensively assess 
the effectiveness of these models in a real-world, multi-dialectal context.

Subtask 2: Cross-dialect Translation and Intent Preservation

Subtask-2 focuses on the translation from MSA to various Arabic dialects: 
Palestinian, Saudi, Tunisian, and Moroccan, with the Palestinian dialect 
provided as a training set. The objective is to retain the original intent in 
the translated dialects, ensuring that the intent detection is as effective as 
the MSA examples. This subtask will test the ability of models to adapt MSA 
banking queries into dialectal Arabic while preserving the semantic integrity, 
crucial for accurate intent classification, despite the potential complexities 
introduced by dialectal variations.
In Subtask-2, participants will receive a parallel dataset containing banking 
queries in both MSA and Palestinian Arabic, and their associated intents. The 
subtask at hand is to accurately translate these queries into the following 
Arabic dialects: Palestinian, Saudi, Tunisian, and Moroccan. Participants will 
use the intent labels to examine if the resulting translations' intents are 
preserved and aligned with the original MSA queries.

For more details please visit: https://sina.birzeit.edu/arbanking77/arafinnlp/

Important Dates

- February 24, 2024: Shared task announcement.
- March 1, 2024: Release of training and development datasets
- April 5, 2024: Registration deadline.
- April 10, 2024: Test set made available.
- April 22, 2024: Codalab Test system submission deadline.
- May 17, 2024: Shared task papers due date
- June 17, 2024: Notification of acceptance
- July 1, 2024: Camera-ready papers due
- August 16, 2024: ArabicNLP conference

Organising Committee

- Mo El-Haj, Lancaster University, United Kingdom
- Houda Bouamor, Carnegie Mellon University, Qatar
- Saad Ezzini, Lancaster University, United Kingdom
- Ismail Berrada, Mohammed VI Polytechnic University, Morocco
- Sanad Malaysha, Birzeit University, Palestine
- Mohammed Khalilia, Birzeit University, Palestine
- Mustafa Jarrar, Birzeit University, Palestine
- Sultan Almujaiwel, King Saud University, Saudi Arabia

Registration

Registration is now open up until 5 April 2024: 
https://forms.gle/MuQb9H1Gq2nQoFUg8



We look forward to your participation in the AraFinNLP shared task.

Best wishes,

AraFinNLP Organising Committee


--------------------------------

Dr Mo El-Haj
Senior Lecturer in NLP
Director of Admissions (SCC)
Co-Director of UCREL NLP Group

Natural Language Engineering (NLE) Journal Editorial Board
https://www.cambridge.org/core/journals/natural-language-engineering


Advisory Board of the Natural Language Processing Book Series
https://benjamins.com/catalog/nlp


School of Computing and Communications, Lancaster University
https://www.lancaster.ac.uk/staff/elhaj
@DocElhaj

You may receive emails from me outside what are your typical office hours.
I do not expect you to respond to my email outside your working hours.

_______________________________________________
Corpora mailing list -- corpora@list.elra.info
https://list.elra.info/mailman3/postorius/lists/corpora.list.elra.info/
To unsubscribe send an email to corpora-le...@list.elra.info

Reply via email to