Frequently Asked Questions

Other information about the ACL Anthology

• How can I make suggestions or provide feedback on the Anthology?

  We welcome your feedback! The most convenient way for us to receive it is for you to create a new issue on our Github issues page. This is also the method most likely to generate the fastest response, since it can be seen by anyone on the volunteer team. If you prefer, you are also invited to email the Anthology director.

• How can I submit corrections to papers?

  There are a number of types of corrections. For an overview, you can see our our corrections page.

  In summary, there are three main types of corrections we handle:

  • Metadata only. This includes corrections to the title, author list, or abstract. These can be initiated by clicking on the yellow “Fix data” button on any paper page in the Anthology.
  • Author disambiguation. These corrections involve author pages (e.g., anything under https://aclanthology/people/), and can also be initiated by clicking on the yellow “Fix data” button on any author page. There are two main types of author disambiguation:
    1. One author, 2+ names. e.g., Aravind Joshi Aravind K. Joshi, etc. Note that often, an author will have multiple pages because the metadata for one or more PDFs is incorrect. The pages can be automatically merged if the underlying PDFs have the verbatim string and are simply correct in the metadata. If different names were used, we have to create an entry in our name variants file.
    2. 2+ authors, one name. e.g., Yang Liu. In this situation, we have to manually assign IDs to the papers to create a separate author page, typically using their Ph.D. granting institution (e.g., Yang Liu of Edinburgh.
  • PDF corrections. This includes changes such as revisions or retractions. For more information, please see see details on the corrections page.

  Important: The Anthology treats the data on the PDF itself as the authoritative source of information, so all metadata corrections should conform to the information found there. If that information is wrong, a PDF correction is needed.

• How do I link to publications in the ACL Anthology?

  Every paper in the Anthology is assigned an Anthology ID. After 2020, identifiers are of the form {year}.{venue}-{volume}.{#}, where {year} is the four-digit year, {venue} a lowercased venue code comprising ASCII letters and digits, {volume} is a volume name or number, and {#} is the paper number. Prior to 2020, this identifier took the form CYY-VPPP or CYY-VVPP, where C is a collection, YY a two-digit year, V a volume, and P a paper ID. The canonical URL of an Anthology paper is given by appending this identifier to the Anthology’s base URL https://www.aclanthology.org/; e.g., https://www.aclanthology.org/2020.iwclul-1.4 (new style) or https://www.aclanthology.org/E91-1001 (old style). This is the paper’s landing page, which includes (among other things) a link to the PDF.

  Many papers in the Anthology also have Digital Object Identifiers (DOIs). Both the DOIs and the canonical Anthology URLs embed the 8-character ACL Anthology Identifier. When available, DOI URLs will redirect to the Anthology canonical URL, and will be listed on that page.

  Variations of the canonical URL can be used to access the PDF and citation format files directly:

  • Append .pdf to get the PDF, e.g., https://www.aclanthology.org/2020.iwclul-1.4.pdf
  • Append .bib to retrieve the BibTeX file, e.g., https://www.aclanthology.org/2020.iwclul-1.4.bib
  • Append .xml to get the MODS-formatted XML file, e.g., https://www.aclanthology.org/2020.iwclul-1.4.xml

  and so on.

• How can I submit attachments to papers?

  To submit paper attachments, use this Microsoft form. Important: If the file attachment cannot be downloaded automatically (via wget), processing of your attachment will fail. Many sharing sites (Github, Google Drive, Dropbox) display by default an HTML page. This complicates processing of attachments and increases the time it takes. Please manually check that you can download your file via wget.

  Attachments are processed at least quarterly, at the beginning of January, April, July, and October. We often also process a week or so after major *ACL conferences. Additional processing is undertaken as volunteer time permits.

• How can I submit my proceedings or journal to the Anthology?

  Please see our submission guide.

• How does the ACL Anthology use Digital Object Identifiers (DOIs)?

  The ACL uses DOIs to refer to scholarly works persistently in accordance with cross-disciplinary standards. Prior to 2012, ACL materials were assigned DOIs by the Association of Computational Machinery (ACM). Since 2014, ACL has assigned its own DOIs to materials using a convention based off of the ACL Anthology Identifiers. In 2015, ACL had to change DOI prefix due to accounting practices for the ACM.

  Prefixes:

  • 2014–2015: 10.3115
  • 2015–current: 10.18653

  We strongly encourage the publication of DOIs in citing bibliographies to help better cross-reference ACL materials.

  Note while the ACL Anthology hosts third-party materials coming from sister societies, these materials are hosted courtesy the ACL but are not assigned DOIs by the ACL due to costs and copyright limitations; the DOI information above is only applicable to ACL sponsored events by ACL, its chapters or SIGs.

• What is the copyright for materials in the ACL Anthology?

  The ACL materials that are hosted in the Anthology are licensed to the general public under a liberal usage policy that allows unlimited reproduction, distribution and hosting of materials on any other website or medium, for non-commercial purposes. Prior to 2016, all ACL materials are licensed using the Creative Commons 3.0 BY-NC-SA (Attribution, Non-Commercial, Share-Alike) license. As of 2016, this policy has been relaxed further, and all subsequent materials are available to the general public on the terms of the Creative Commons 4.0 BY (Attribution) license; this means both commercial and non-commercial use is explicitly licensed to all.

  Note that these policies only cover ACL materials. As with the DOIs, this policy does not cover third-party materials. For reproduction privileges for such materials, please approach the respective organizations.

• How can I help improve the ACL Anthology?

  We welcome your contributions! In fact, since the Anthology is run entirely on volunteer energy, we need them. The best way to contribute is to browse our Github issues page (we suggest starting with the “help wanted” tag, and see what needs to be done that overlaps with your interests or skills. Please also feel free to email the director, anthology@aclweb.org. You might also be interested in our developer documentation.

• What other resources are there based on the ACL Anthology?

  The following papers provide overviews of the Anthology infrastructure at different points in time:

  • Two Decades of the ACL Anthology: Development, Impact, and Open Challenges (Bollmann et al., NLPOSS-WS 2023).
  • The ACL Anthology: Current State and Future Directions (Gildea et al., NLPOSS 2018). It is now out of date, but helped direct changes to the Anthology that occurred in 2019.

  There are also many papers that make use of the Anthology’s data. Here we highlight a few:

  • On Forgetting to Cite Older Papers: An Analysis of the ACL Anthology (Bollmann & Elliott, ACL 2020)
  • NLP Scholar: An Interactive Visual Explorer for Natural Language Processing Literature (Mohammad, ACL 2020)
  • Mining Scientific Terms and their Definitions: A Study of the ACL Anthology (Jin et al., EMNLP 2013)
  • The ACL Anthology Reference Corpus: A Reference Dataset for Bibliographic Research in Computational Linguistics (Bird et al., LREC 2008). This work provided plain text and other metadata for a controlled subset of the ACL Anthology.

  Many other publications based on the ACL Anthology exist, which you may be able to find by searching.

• What software does the ACL Anthology use?

  The ACL Anthology is built using completely open-source software, with further developments and customization done by the ACL Anthology staff. The current Anthology website is statically generated using the Hugo framework, with the layout being designed using the Bootstrap library. Font Awesome is used as the icon font.

  The source code for the Anthology is available on GitHub at https://github.com/acl-org/acl-anthology/.

• How can I programmatically access the Anthology's data?

  Please see our developer documentation.