Subject: Session Summary – Universal Acceptance Day 2026: Bengali & Urdu Localization Project

 

Dear NARALO Group,

 

I hope this email finds you well.

 

Following Glenn McKnight's message sharing the recordings, I wanted to specifically highlight the "UA and Education" presentation, in which I participated. For those who weren't able to attend or want a reference summary, here's what was covered.

 

 

-----

SESSION OVERVIEW

 

The session focused on a volunteer-led initiative to localize a 10-module internet governance curriculum — developed through VSIG (Virtual School of Internet Governance) — into Bengali and Urdu, with the goal of making foundational internet governance education accessible to over 300 million Bengali speakers and hundreds of millions of Urdu speakers.

 

-----

BENGALI LOCALIZATION – KEY HIGHLIGHTS (Presented by Mohibul Mahmud & Zulqar Nayen)

 

Project Timeline: The project ran from November 2024 through the UA Day 2026 presentation, progressing through team formation, translation, iterative review, evaluator assessment, and a live pilot at the University of Dhaka.

 

Methodology: The team used an AI-human hybrid workflow: AI generated first drafts at speed; human reviewers then ensured accuracy, natural flow, and educational clarity. A three-part terminology strategy was adopted — translate, transliterate, or retain English — depending on what best served comprehension for Bengali learners.

 

University of Dhaka Pilot:

- 38 students enrolled from the Department of Management Information Systems.

- 18 students (47%) completed all modules and passed all quizzes.

- For context, typical MOOCs globally see 5–15% completion rates.

- Dr. Rukibul (Head, MIS Department) provided institutional credibility and himself completed the course.

- An in-person visit to Dhaka in October 2025 was identified as a key turning point in establishing the partnership.

 

Universal Acceptance Findings: The pilot surfaced real UA and internationalization failures that were not anticipated:

- Email Address Internationalization (EAI): Automated Moodle emails silently failed to deliver when encountering internationalized address configurations, with no error logs generated.

- Bengali script rendering: Conjunct characters (multiple consonants merging into a single visual form) were stored correctly in the database but stripped at the browser display layer, altering the meaning of words.

- The team noted that platforms failing to render Bengali script in a paragraph are almost certainly going to fail at accepting Bengali-script domain names — making localization projects a practical UA readiness audit.

 

Lessons Learned: What worked well: pre-translation style and terminology discussions; the AI-human workflow; the in-person Dhaka meeting; student feedback as an unplanned QA resource. What to improve next time: establish a comprehensive shared glossary before translation begins; implement structured version tracking; conduct UI/UX and Unicode compliance testing before the pilot; formalize student feedback mechanisms.

 

Five-Pillar Replication Framework: The team distilled their experience into a playbook for other language communities:

1. Terminology Kit – shared glossary and decision rules before work begins

2. AI Governance – structured AI-human hybrid workflow

3. Script Engineering – verify Unicode rendering and font fallback for non-Latin scripts early

4. Academic Anchor – partner with a university for institutional credibility and accountability

5. Expanding Spiral – start small, pilot, gather data, refine, and repeat

 

-----

URDU LOCALIZATION – KEY HIGHLIGHTS (Presented by Waqar Ahmad)

 

The Urdu team is currently in progress, slightly behind the Bengali project. Plans include a pilot in Pakistan in collaboration with the Internet Society's local chapter and the Virtual University of Pakistan. Future language expansions being explored include Hindi, Swahili, and Arabic.

 

-----

FUNDING & SUSTAINABILITY

 

The project was almost entirely volunteer-driven. Total funding received was approximately $2,000 across both the Bengali and Urdu efforts — a nominal stipend distributed to contributors on the ground. The team noted that internet governance education does not fit neatly within any single government ministry's mandate in most developing countries, making grassroots coordination the only viable path for now.

 

-----

CLOSING REFLECTIONS

 

Speakers shared what they were most proud of:

- Zulqar Nayen: proving that a small volunteer team with no meaningful budget can localize a university-level curriculum using AI tools and deploy it at a top institution within a year — and documenting the process so others can replicate it.

- Mohibul Mahmud: the team's follow-through on a commitment made early in the project, and the hope that even a small pilot in Dhaka could inspire others to engage with internet governance in their own language.

 

Best regards,

Mohibul Mahmud

NARALO Member