Workshop for Industrial NLP
17 October 2026 · Co-located with INLG 2026 · Half-Day Workshop
Welcome to the first Workshop for Industrial NLP, a half-day workshop at INLG 2026 dedicated to natural language generation (NLG) and large language model (LLM) research shaped by real deployment settings.
By industrial NLP, we mean NLP and NLG research driven by the realities of production: deployment constraints, domain-specific risks, user-facing applications, business and public-sector workflows, and the need for measurable reliability beyond benchmark performance.
Recent progress in generative NLP has accelerated industrial adoption across customer support, content generation, finance, cybersecurity, education, healthcare, software engineering, and multilingual communication. Yet deployed systems expose limitations that standard benchmarks often miss: hallucination, inconsistent behavior, prompt sensitivity, safety failures, bias, privacy risks, insufficient domain adaptation, and weak evaluation protocols. This workshop brings together researchers and practitioners working on NLG, LLMs, evaluation, safety, fairness, multilingual NLP, and deployed applications, with the goal of supporting industrial systems that are reliable, accountable, robust, and useful in real-world environments.
🔗 Read the full Call for Papers →
Objectives
- Create a focused venue for NLG work on safety, reliability, evaluation, and industrial deployment.
- Encourage dialogue between academic researchers and industry practitioners about practical failure modes and realistic evaluation needs.
- Highlight methods for making generated text more controllable, trustworthy, fair, multilingual, and domain-robust.
- Identify open research directions where INLG can contribute to responsible and high-impact NLP systems.
Important Dates
| Milestone | Date |
|---|---|
| Direct submission deadline | 16 August 2026 |
| ARR commitment deadline | 29 August 2026 |
| Notification of acceptance | 9 September 2026 |
| Camera-ready due | 16 September 2026 |
| Workshop | 17 October 2026 |
All deadlines follow the INLG suggested schedule. Times are end-of-day (anywhere on Earth) unless otherwise noted.
Call for Papers
We solicit short papers, extended abstracts, position papers, and industrial case studies. Authors may choose archival or non-archival submission, making the workshop suitable both for mature research and for emerging industrial case studies.
Submissions are invited under seven tracks:
- Safety & Security
- Reliability & Fairness
- Efficiency
- Industrial Applications & Deployment (e.g. medicine, finance, cybersecurity)
- Evaluation, Datasets & Benchmarks
- NLP for Social Good & Real-World Impact
- Negative Impacts & Their Mitigations
🔗 See the full Call for Papers, tracks, and submission details →
Program
The workshop takes place on 17 October 2026, 13:30–17:00, combining keynotes, paper presentations, and discussion. The program is:
| Time | Session | Duration | Format |
|---|---|---|---|
| 13:30–13:35 | Opening | 5 min | Welcome and introduction. |
| 13:35–14:15 | Keynote 1: Qing Li | 40 min | 30 min talk + 10 min Q&A. |
| 14:15–15:00 | Papers 1–3 | 45 min | Each paper: 12 min talk + 3 min Q&A. |
| 15:00–15:25 | Coffee break | 25 min | Informal discussion. |
| 15:25–16:05 | Keynote 2: Gerald Penn | 40 min | Invited talk. |
| 16:05–16:50 | Papers 4–6 | 45 min | Each paper: 12 min talk + 3 min Q&A. |
| 16:50–17:00 | Closing remarks | 10 min | Summary and takeaways. |
| 17:00 | End | — | — |
Invited Speakers
- Qing Li (University of Groningen) — Keynote 1
- Gerald Penn (University of Toronto) — Keynote 2
Organizers
- Yizhi Li — University of Manchester (reasoning and CLI agents)
- Yi Zhou — Cardiff University (responsible AI)
- Jingcheng Niu — TU Darmstadt (mechanistic interpretability)
- Haau-Sing Li — TU Darmstadt and Instituto de Telecomunicações (code generation and test-time compute)
- Joanne Boisson — Cardiff University (social NLP and metaphor)
- Yi Qi — University of Leeds (safety assurance and trustworthy AI)
- Jie Fu — IQuest Research (safety)
Contact
Updates, submission instructions, and the submission portal will be posted on this website. The call is circulated through SIGGEN, ACL mailing lists, social media, and relevant academic and industry networks.
We look forward to your submissions and to a focused discussion of how NLG research can shape the next generation of deployed NLP systems.