.png)
SilkRoadNLP 2027
Call for Papers
Advaning NLP for the Iranian Language Family
March 9-14, 2027. Athens, Greece.
2027 Theme
From Representation to Sovereignty: AI for Iranian Languages
co-located with
Paper Submission Deadline: 15 December 2026
Overview
SilkRoadNLP 2027 provides a platform for advancing Natural Language Processing (NLP), computational linguistics, and AI for the Iranian linguistic family—a diverse group of languages spoken across Iran, Afghanistan, Central and South Asia, and the Caucasus, as well as diaspora communities worldwide. We welcome work that bridges computational methods with linguistic, social, and cultural perspectives.
Despite major advances in multilingual NLP and large language models, the languages of the Iranian linguistic group remain underrepresented in mainstream research. While Persian has received some attention, many languages and varieties of the Iranian language family such as Kurdish, Balochi, Gilaki, Mazandarani, Luri, Pashto, Ossetic, Talysh, Zazaki, and others, continue to have limited digital presence, computational resources, benchmarks, and system support. Many are minority or endangered languages at risk of further technological marginalization. Within Persian, Iranian Persian, Dari, and Tajik differ in orthography, phonology, lexicon, and sociolinguistic context, with Tajik also differing in script.
Building on the inaugural workshop at EACL 2026, SilkRoadNLP continues its mission as a dedicated ACL-affiliated venue for computational research on the Iranian language family. The workshop aims to:
-
Highlight advances in NLP analysis, datasets, benchmarks, models, and language technologies.
-
Increase the visibility of Iranian languages and varieties underrepresented in mainstream NLP and AI.
-
Address linguistic, cultural, sociopragmatic, and representational challenges in computational research.
-
Foster collaboration across computational linguistics, linguistics, digital humanities, language documentation, education, and AI governance.
-
Build a sustained international research community and shared ecosystem of resources for Iranian-language NLP.
We welcome research on all languages of the Iranian linguistic family, including contemporary, historical, regional, minority, endangered, heritage, and diaspora varieties.
2027 Theme: From Representation to Sovereignty: AI for Iranian Languages
This year’s theme moves beyond asking whether Iranian languages are represented in NLP and AI to examine how they are represented, who participates in developing the technologies around them, who governs linguistic and cultural data, and how language communities can retain meaningful agency over AI systems.
We especially welcome work on community-centered NLP, culturally grounded evaluation, data and language sovereignty, participatory development, and responsible AI across the development lifecycle. The workshop also aims to lay the groundwork for a collaborative Sovereign AI agenda for the Iranian linguistic family.
The theme complements the workshop’s broad scope. Contributions across Iranian-language NLP are welcome, whether or not they directly address the annual theme.
Topics of Interest
We invite submissions on any aspect of NLP, computational linguistics, or AI relevant to the Iranian linguistic family, including contemporary, historical, regional, minority, endangered, heritage, and diaspora varieties. Topics include, but are not limited to:
-
Computational linguistic analysis of Iranian languages: morphology, syntax, semantics, pragmatics, and typology.
-
Resource creation, including corpora, benchmarks, and lexicons; language documentation and annotation.
-
Speech and multimodal processing, including automatic speech recognition, text-to-speech, and multimodal learning.
-
OCR, digitization tools, and script technologies.
-
Language modeling and LLM adaptation, fine-tuning, and evaluation across domains.
-
NLP for historical and extinct Iranian languages, such as Pahlavi, Avestan, and Sogdian, including applications in digital humanities and cultural heritage.
-
NLP for regional varieties, dialects, heritage languages, and minority and endangered languages.
-
Machine translation, transliteration, and cross-script processing.
-
Evaluation frameworks, benchmarks, and shared tasks.
-
Educational applications and language learning tools.
-
Culturally grounded and sociopragmatic evaluation of language technologies.
-
Community-centered NLP, participatory development, and collaboration with language communities.
-
Data and language sovereignty, community agency, and governance of linguistic and cultural resources.
-
Ethical and responsible AI across the language technology development lifecycle.
-
Downstream and applied NLP, including information extraction, sentiment analysis, misinformation detection, information retrieval, social media analytics, and health and legal applications.
Contributions need not directly address the annual theme; we welcome work across the full range of Iranian-language NLP.
Shared Task: Tajik Cyrillic <-> Perso-Arabic Transliteration
SilkRoadNLP 2027 will introduce a shared task on transliteration between Tajik Cyrillic and Perso-Arabic scripts. Building on an existing aligned dataset and established evaluation metrics, the task will provide a common benchmark for comparing transliteration systems and establishing reproducible baselines and state-of-the-art performance.
The shared task aims to advance tool development and standardization for cross-script processing within the Iranian language family.
Task specifications, data release and participation dates, evaluation procedures, and system paper submission instructions are being finalized and will be announced on the workshop website.
Paper Submission Format
We invite the following categories of submissions:
-
Long papers: Up to 8 pages of content (excluding references). Long papers should present substantial, original, unpublished contributions, with appropriate evaluation and analysis.
-
Short papers / demo papers: Up to 4 pages of content (excluding references). Short papers should present a focused contribution that can be adequately described within this length. Demo papers should describe a system or tool and its relevance to the workshop.
We welcome research, position papers, resource papers, system demonstrations, and work-in-progress contributions within these categories.
All papers submitted to the workshop must be fully anonymized for double-blind review.
All submissions must use the official ACL style template and be submitted as PDF files. References, appendices, a dedicated Limitations section, and an optional Ethical Considerations section do not count toward the content page limit. Appendices should appear after the references. The main paper must contain the information necessary to assess the contribution; reviewers are not required to read supplementary materials.
Submission & Review
Direct submissions will be made through the workshop’s OpenReview submission page. We will also accept papers with completed ACL Rolling Review (ARR) reviews, including a meta-review. Instructions for submitting ARR-reviewed papers and the workshop submission link will be announced here.
Camera-ready instructions will be provided with acceptance notifications. Accepted papers will be published in the ACL Anthology as part of the EACL 2027 workshop proceedings.
Hybrid Participation
The workshop will be held in a collaborative hybrid format, combining in-person sessions in Athens with online participation to ensure wide accessibility. The workshop will feature live-streamed talks, virtual poster sessions, and interactive Q&A opportunities for remote attendees. Final details on the hybrid setup will be shared closer to the event in coordination with EACL 2027.
Important Dates
-
Call for Papers: October 8, 2026
-
Direct Submission Deadline: December 15, 2026
-
Pre-reviewed ARR Commitment Deadline: December 22, 2026
-
Notification of Acceptance: January 5, 2027
-
Camera-ready Papers Due: January 9, 2027
-
Poster Boards and presentation slides Due: February 5, 2027
-
Workshop Date: TBA / March 9-14, 2027
All deadlines are 11:59 PM UTC-12 (“anywhere on Earth”).
Diversity & Inclusion
SilkRoadNLP is committed to inclusive participation. We strongly encourage submissions from early-career researchers, independent scholars, and institutions in low-resource regions. The workshop will support hybrid participation (onsite + virtual) to maximize accessibility.
Questions?
Feel free to contact us with any questions.