The 2nd Call: December 20, 2022
ACL 2023 invites the submission of long and short papers featuring substantial, original, and unpublished research in all aspects of Computational Linguistics and Natural Language Processing. As in recent years, some of the presentations at the conference will be of papers accepted by the Transactions of the ACL (TACL) and by the Computational Linguistics (CL) journals.
ARR commitment and Direct submission: https://softconf.com/acl2023/papers/
Major differences from “standard” recent conferences include:
- Hybrid ARR + ACL models.
- The mandatory section of “Limitations”
- Responsible NLP Research checklist
- Theme track: “Reality Check”.
All deadlines are 11:59PM UTC-12:00 (“anywhere on Earth”).
|Submission template available
|November 1, 2022
|Anonymity period for ARR papers
|November 15, 2022
|Submission deadline for papers submitted to ARR
|December 15, 2022
|Anonymity period for papers submitted through Softconf START
|December 20, 2022
|Abstract deadline for Softconf START direct submissions
|January 13, 2023
|Direct paper submission deadline
|January 20, 2023
|Anonymity period for ARR papers to be committed to ACL 2023
|Februrary 17, 2023
|Commitment deadline for ARR papers
|March 17, 2023
|Author response period
|March 17-24, 2023
|March 25-April 7, 2023
|Notification of acceptance
|May 1, 2023
|May 8, 2023
|Camera-ready papers due
|May 25, 2023
|July 9, 2023
|July 10-12, 2023
|July 13-14, 2023
Mandatory abstract submission
The paper title, author names, contact details, and a brief abstract must be submitted electronically through the ACL 2023 paper submission site by the abstract submission deadline (January 13). It will be possible to make minor edits to the title and abstract until the full paper submission deadline, but you cannot change authors and subject areas. Submissions with “placeholder” abstracts will be removed without consideration.
Important: if you miss the abstract submission deadline, then you cannot submit the full paper.
ACL 2023 aims to have a broad technical program. Relevant topics for the conference include, but are not limited to, the following areas (in alphabetical order):
- Computational Social Science and Cultural Analytics
- Dialogue and Interactive Systems
- Discourse and Pragmatics
- Ethics and NLP
- Information Extraction
- Information Retrieval and Text Mining
- Interpretability and Analysis of Models for NLP
- Language Grounding to Vision, Robotics and Beyond
- Large Language Models
- Linguistic Diversity
- Linguistic Theories, Cognitive Modeling, and Psycholinguistics
- Machine Learning for NLP
- Machine Translation
- Multilingualism and Cross-Lingual NLP
- NLP Applications
- Phonology, Morphology, and Word Segmentation
- Question Answering
- Resources and Evaluation
- Semantics: Lexical
- Semantics: Sentence-level Semantics, Textual Inference, and Other Areas
- Sentiment Analysis, Stylistic Analysis, and Argument Mining
- Speech and Multimodality
- Syntax: Tagging, Chunking and Parsing
- Theme Track (see below)
Theme Track: Reality Check
Following the success of the ACL 2020-2022 Theme tracks, we are happy to announce that ACL 2023 will have a new theme with the goal of reflecting and stimulating discussion about the current state of development of the field of NLP. While the current systems perform much better and fail more gracefully than their rule-based predecessors, there are growing piles of evidence of other kinds of brittleness, including out-of-domain generalization, adversarial attacks, spurious patterns (both linguistic and social), lack of sensitivity to basic linguistic perturbations such as negation, over-sensitivity to perturbations that should not matter (e.g. order and wording of prompts), etc.
The theme track invites empirical and theoretical research, as well as position and survey papers reflecting on the ways in which reported performance improvements on NLP benchmarks are meaningful. The possible topics of discussion include (but are not limited to) the following:
- How reliably do the leaderboard scores translate to improvements in real-world use of the models?
- How reliably do the leaderboard scores compare competing models?
- While the current NLP systems are not brittle in the same way as their predecessors, they are still brittle in other ways. What tasks can we claim to have “solved”, if any?
- Have performance improvements been accompanied by commensurate growth in the scientific understanding (of language, cognition, or deep learning technology)? In what ways?
- Given that the authors of engineering papers are incentivized to report only the most successful results, especially for the systems that are also commercial products, what can the NLP venues do to improve reporting?
The theme track submissions can be either long or short. We anticipate having a special session for this theme at the conference and a Thematic Paper Award in addition to other categories of awards.
Paper Types and Formats
Long papers must describe substantial, original, completed, and unpublished work. Wherever appropriate, concrete evaluation and analysis should be included. Long papers may consist of up to eight (8) pages of content, plus unlimited pages of references and appendices. Final versions of long papers will be given one additional page of content (up to nine (9) pages), so that reviewers’ comments can be taken into account.
Short paper submissions must describe original and unpublished work. Please note that a short paper is not a shortened long paper. Instead, short papers should have a small, focused contribution - a point that can be made in a few pages with sufficient level of detail. Short papers may consist of up to four (4) pages of content, plus unlimited pages of references and appendices. Final versions of short papers will be given one additional page of content (up to five (5) pages), so that reviewers’ comments can be taken into account.
Both short and long papers are submitted through the same process, but they will be reviewed using a different review form and criteria. Short and long papers will be distinguished in the proceedings, but the length of the paper does not correspond to either the final presentation format (poster or oral), or the type of contribution made in the paper. In particular, ACL 2023 welcomes the following kinds of contributions:
- Computationally-aided linguistic analysis (of either models or data resources)
- NLP engineering experiment
- Reproduction study
- New data resources, particularly for low-resource languages
- Approaches for data- and compute efficiency
- Position papers
- Publicly available software and pre-trained models
While there is no direct mapping between types of contributions and paper length, some kinds of papers naturally gravitate towards a certain length: e.g. surveys are more likely to be long rather than short papers. One paper can make more than one contribution of different types.
Paper Submission and Templates
Submission is electronic, using the Softconf START conference management system. Both long and short papers must follow the ACL 2023 two-column format, using the supplied official style files. The templates can be downloaded in Style Files and Formatting. Please do not modify these style files, nor should you use templates designed for other conferences. Submissions that do not conform to the required styles, including paper size, margin width, and font size restrictions, will be rejected without review.
To guarantee conformance to publication standards, we will be using the ACL Pubcheck tool. The PDFs of camera-ready papers must be run through this tool prior to their final submission, and we recommend its use also at submission time.
Paper Submission Process
Following EMNLP 2022 and EACL 2023, ACL 2023 will continue to offer a hybrid submission format with respect to ACL Rolling Review (ARR): the submissions may come through ARR, but it will also be possible to submit directly to ACL through the Softconf START system. A given submission may only choose one of these two options, i.e. it is not possible to submit a paper to ACL directly while being under review at ARR and vice versa.
The papers submitted directly to ACL will have the “regular” review process: a paper is reviewed by 3 reviewers, and authors are invited to write an author response and revise their paper before the camera-ready deadline, if accepted. The deadline for submitting papers through the START system is January 13, 2023.
Papers submitted to ACL 2023, but not selected for the main conference, will also automatically be considered for publication in the Findings of the Association of Computational Linguistics. The notifications for the main track and Findings will come out simultaneously, and the authors of submissions accepted to Findings may opt to withdraw until May 8.
ARR papers committed to ACL will be handled by the Senior Area Chairs. For these papers, the authors may provide an author response but not revise their paper (with the exception of adding the required “Limitations” section, if it is missing from the ARR submission; see below).
To be considered by ACL 2023, ARR papers need to be submitted to ARR and obtain all its reviews and meta-review by the commitment deadline of March 17. These papers cannot be modified except that they can be associated with an author response. They also cannot be revised and re-submitted to ARR for a new review cycle while they are considered by ACL 2023.
Note: Generally, to receive reviews by the commitment deadline, the regular ARR timeline suggests that ARR suggestions should be made by December 15th. Note that the cycle length will change in December to 8 weeks.
If you have submitted a paper to ARR but do not want to use the review result of ARR, it must be inactive in the ARR system in order for the paper to be submitted directly to ACL. This means that the submission must either be explicitly withdrawn by the authors, or the ARR reviews are finished and shared with the authors before January 1st, 2023, and the paper was not re-submitted to ARR. Authors can withdraw a paper from ARR by January 1st, regardless of how many reviews they have received, and submit the paper directly to ACL 2023.
Papers submitted to ACL 2023 may not be submitted for review elsewhere (including ARR) while being under review at ACL 2023. However, we allow an exception of the multiple submission policy for papers under review for EACL 2023 and ICLR 2023. See the Multiple Submission Policy below.
Paper Submission Policies
All submissions to ACL 2023 must follow the following policies.
The direct submissions to ACL may not be made available online (e.g. via a preprint server) in a non-anonymized form after December 20 11:59PM UTC-12:00 (for arXiv, note that this refers to
publication time, not submission time). The respective deadline for ARR is November 15th. Non-anonymous preprints that were published before the start of the anonymity period may not be updated until ACL notifications come out. The only exception is for the purpose of correcting names, in which case the PC chairs should be notified per ACL policy. The existence of non-anonymous preprints must be disclosed in the submission form.
ACL does not prohibit giving talks about work under review in small groups, but we ask you not to advertise the preprint (or such talks) on social media, blog about this work, or have it covered in the media during the anonymity period. Anonymous preprints (e.g. on ARR) can be posted after the start of the anonymity period, but likewise should not be advertised by their authors or their close colleagues, as that can compromise the review process.
Anonymization for Double-blind Review
The submissions to ACL and any supplementary materials must not include authors’ names and affiliations, or the acknowledgments section. Self-references that reveal the authors’ identities must be avoided. For example, instead of “We previously showed (Smith, 1991) …” or even “We previously showed (Anonymous, 1991)… “ please use “Smith previously showed (Smith, 1991) …”.
The submissions should avoid links to non-anonymized repositories: the code should be either submitted as supplementary material, or as a link to an anonymized repository (e.g., Anonymous GitHub or Anonym Share). Please avoid any links to storage services like Dropbox (which may track the reviewers downloading the resources). Papers that do not conform to these requirements will be rejected without review.
Citation and Comparison
See the ACL guidelines.
The author list for submissions should include all (and only) individuals who made substantial contributions to the work presented. The list of the authors (including the order) may not be changed after submission.
Optional Supplementary Materials (Appendices, Software, Data)
See the policy in the ARR CFP. Papers should not refer for further detail, to documents that are not available to the reviewers. If important citations are not available to reviewers (e.g., awaiting publication), these paper/s should be anonymised and included in the appendix. They can then be referenced from the submission without compromising anonymity.
Both long and short papers must follow the ACL 2023 two-column format, using the supplied official style files. The templates can be downloaded in Style Files and Formatting. Please do not modify these style files, nor should you use templates designed for other conferences. Submissions that do not conform to the required styles, including paper size, margin width, and font size restrictions, will be rejected without review. We recommend using the ACL Pubcheck tool to confirm that your submission conforms to the standards prior to submission, so as to avoid desk rejections.
Mandatory discussion of limitations
We believe that it is also important to discuss the limitations of your work, in addition to its strengths. ACL 2023 requires all papers to have a clear discussion of limitations, in a dedicated section titled “Limitations”. This section will appear at the end of the paper, after the discussion/conclusions section and before the references, and will not count towards the page limit. Papers without a limitation section will be automatically rejected without review.
An ARR-reviewed paper that did not include “Limitations” section in their prior submission, should submit a PDF with such a section together with their ACL 2023 submission. We treat ARR papers with and without the Limitations section equally if the Limitation section is provided in the paper or as an additional PDF.
While we are open to different types of limitations, just mentioning that a set of results have been shown for English only probably does not reflect what we expect. Mentioning that the method works mostly for languages with limited morphology, like English, is a much better alternative. In addition, limitations such as low scalability to long text, the requirement of large GPU resources, or other things that inspire crucial further investigation are welcome.
Scientific work published at ACL 2023 must comply with the ACL Code of Ethics. We encourage all authors to include an explicit ethics statement on the broader impact of the work, or other ethical considerations after the conclusion but before the references. The ethics statement will not count toward the page limit (8 pages for long, 4 pages for short papers).
Responsible NLP Checklist
ACL 2023 submissions are required to fill the Responsible NLP checklist in the paper submission form (not as a supplementary pdf file, like before) that covers reproducibility and ethical considerations. We encourage the authors to have a look at the Responsible NLP checklist as early as possible in the cycle of their project, to make sure that any necessary IRB permissions are obtained, experiment parameters are saved, and energy expenditure is estimated. The checklist will include one new question asking to disclose the use of AI writing assistants, if any.
Multiple Submission Policy
In principle, papers submitted directly to ACL must not be under review for any other archival venue at any time during the ACL review process. We will not consider any paper that overlaps significantly in content or results with papers that will be (or have been) published elsewhere. However, considering the overlap of the paper submission due and the notification of acceptance of EACL 2023 and ICLR 2023, we allow an exception for the multiple submission policy; a work under review for EACL 2023 or ICLR 2023 can be submitted to ACL, but must be declared at submission time. If the work is accepted, it must be withdrawn from ACL no later than January 31 (when accepted to EACL 2023) and no later than January 22 (when accepted to ICLR 2023). Papers submitted to ICLR 2023 and ACL 2023 must follow the anonymization policy of ACL 2023.
Optional: Sticky Reviews
The papers previously reviewed at other *ACL venues (but not through ARR) have the option to submit the paper together with the link to the previous Softconf START submission, from which the track chairs will be able to access the old reviews. They will also be able to submit a short explanation of how the paper was changed in response to the old reviews. This option could be beneficial for the authors who have addressed the problems identified before, and can argue strongly for how the paper has been improved. The prior reviews will not be seen by the new reviewers, but they may be used by the area chairs and program chairs in review quality control, resolving disagreements between reviewers, and in deciding borderline papers.
Cross-submissions with ACL Workshops
The papers submitted for reviewing directly to ACL 2023 and not selected by the main conference will be able to submit to ACL 2023 workshops after the ACL notifications, with the reviews and meta-reviews they received at ACL. Papers accepted to Findings can also apply for presentation at participating workshops. The deadlines for direct submission to be reviewed by the workshops will be determined by the individual workshops.
Authors are required to honor the ethical code set out in the ACL Code of Ethics.
As mentioned above, ACL 2023 will incorporate the Responsible NLP Research checklist into the reviewing process. It should accompany all submissions, and its goal is to present structured information about the ethics and reproducibility aspects of the study. It will be available for the reviewers, as well as any future readers of the paper. Especially for the junior authors, we strongly suggest that they consider the checklist at the early stages of their project, rather than treating such questions as an afterthought.
Should the reviewers still have concerns about the ethical aspects of a given submission, it may be passed on to the ACL ethics committee for extra ethics review. ACL reserves the right to reject papers on ethical grounds, where the authors are judged to have operated counter to the code of ethics, or have inadequately addressed legitimate ethical concerns with their work.
All papers accepted to the main conference track must be presented at the conference to appear in the proceedings, and at least one author must register for ACL 2023. All papers accepted to the main track will be required to submit a presentation video. Findings papers may present a virtual poster, a short in-person spotlight, and upload a video (although they are not required to do so). The conference will be hybrid, with an emphasis on encouraging interaction between the online and in-person modalities, and thus presentations can be either on-site or virtual.
- Jordan Lee Boyd-Graber, University of Maryland (USA)
- Naoaki Okazaki, Tokyo Institute of Technology (Japan)
- Anna Rogers, University of Copenhagen (Denmark)