Findings of the Association for Computational Linguistics: EMNLP 2024

Yaser Al-Onaizan, Mohit Bansal, Yun-Nung ChenEMNLP 2024

TL;DR

This document presents the findings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), highlighting the significant growth in submissions and participation. It details the conference's review process, acceptance rates, special themes, and program composition, showcasing the vibrant and evolving landscape of NLP research.

What Problem Does This Paper Address?

The paper addresses the challenge of organizing and presenting the vast amount of research generated by the NLP community at a major conference like EMNLP. It aims to provide a comprehensive overview of the conference's proceedings, including the scale of submissions, the rigor of the review process, and the key research areas and trends that emerged. This serves as a record and a snapshot of the state of the art in empirical NLP.

Methods

The paper describes the two-stage review process, which involves ACL Rolling Review (ARR) followed by a commitment site for EMNLP. It also details the program composition, including oral and poster presentations, and the inclusion of papers from affiliated journals like Computational Linguistics and TACL. The paper highlights the use of a hybrid format (in-person and remote) to maximize accessibility.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Exponential Growth: EMNLP 2024 marks the 29th edition, with a significant increase in papers and registrations compared to its early years, indicating the field's rapid expansion.
  • Hybrid Format Success: The hybrid format (in-person and remote) is a key organizational choice, aiming to accommodate a wider range of participants and streamline the conference experience.
  • Rigorous Review Process: The two-stage review process, leveraging ARR and a dedicated EMNLP commitment site, ensures thorough evaluation of submissions.
  • Special Theme Focus: The "Efficiency in Model Algorithms, Training, and Inference" special theme highlights a critical research direction in NLP, focusing on making models more scalable, practical, and resource-efficient.
  • Commitment to Inclusivity: The emphasis on Diversity & Inclusion chairs and initiatives like the ACL Fellows Lunch demonstrates a commitment to fostering a more welcoming and equitable research community.

Key Results

  • Acceptance Rates: The Main Conference papers had an acceptance rate of 20.8%, with 1271 papers accepted. Findings papers had an acceptance rate of 16.9%, with 1029 papers accepted.
  • Submission Volume: Over 3,500 registrations were recorded, with a significant portion attending in person.
  • Special Theme Submissions: The "Efficiency" track received 138 submissions, with 54 accepted to the main conference and 42 to Findings.

Limitations and Future Work

The paper itself is a report on the conference findings and does not present limitations of a specific research study. However, the scale of the conference and the increasing complexity of organizing such events suggest ongoing challenges in managing logistics and ensuring accessibility for a global community. The continuous growth also implies a need for evolving review processes and presentation formats to accommodate the expanding volume of high-quality research.

Summary updated

BibTeX
@inproceedings{alonaizan2024findings,
  author = {Yaser Al-Onaizan and Mohit Bansal and Yun-Nung Chen},
  title = {{Findings of the Association for Computational Linguistics: EMNLP 2024}},
  booktitle = {Conference on Empirical Methods in Natural Language Processing},
  year = {2024},
}
Figure