Wikipedia Translation Guidelines for LLMs and Machine Translation
Translating content from one language to another is a vital part of expanding the global knowledge base of English Wikipedia. With the rise of Large Language Models (LLMs) and advanced machine translation tools, the process has become faster, but it requires a disciplined approach to maintain encyclopedic standards.
The Role of Machine Translation
Tools such as DeepL and Google Translate serve as excellent starting points for translating articles. However, these tools are not substitutes for human oversight. Translators are expected to follow specific LLM translation guidelines, revising errors and confirming the accuracy of the text rather than simply copy-pasting machine-generated output into the encyclopedia.
[ไม่มีภาพประกอบ]
Quality Control and Verification
Not all source material is suitable for translation. Editors should avoid translating text that appears to be low-quality or unreliable. To ensure the integrity of the English Wikipedia, it is essential to verify the claims in the source text using the references provided in the original foreign-language article.
Categorization and Organization
Proper categorization helps users find content more efficiently. With over 1,053 articles already existing in the main category, editors are encouraged to specify a topic using the | topic = parameter in the translation template to aid in precise categorization.
Copyright and Attribution Requirements
Because Wikipedia content is collaborative, proper credit must be given to the original authors of the source text. This is handled through copyright attribution in the edit summary of the translation.
A valid attribution must include an interlanguage link to the source article. For example, a model edit summary would read: "Content in this edit is translated from the existing Spanish Wikipedia article at [[:es:Corporación Eléctrica Nacional]]; see its history for attribution."
Additionally, editors may place the {{Translated|es|Corporación Eléctrica Nacional}} template on the article's talk page to further document the translation process.
Key Facts
- Machine translation (e.g., Google Translate, DeepL) must be revised and verified, not copy-pasted.
- Unreliable or low-quality source text should not be translated.
- Verification should be performed using the references found in the original article.
- Copyright attribution is mandatory in the edit summary via interlanguage links.
- The
{{Translated}}template can be used on talk pages for documentation.
| Requirement | Action Required | Purpose |
|---|---|---|
| Tool Usage | Revise LLM/Machine output | Ensure accuracy |
| Source Quality | Verify with references | Prevent low-quality content |
| Attribution | Interlanguage link in edit summary | Copyright compliance |
| Categorization | Specify | topic = |
Improve organization |
Frequently Asked Questions
Can I copy and paste directly from Google Translate?
No. Machine translation should only be used as a starting point. You must revise the text for errors and confirm its accuracy before adding it to English Wikipedia.
What should I do if the source text seems unreliable?
You should not translate text that appears to be low-quality or unreliable. Use the references provided in the foreign-language article to verify the information first.
How do I provide proper copyright attribution?
You must include an interlanguage link to the source article in your edit summary, directing users to the source's history for full attribution.
Where can I find more detailed guidance on this process?
For comprehensive instructions, you can refer to the Wikipedia:Translation page.
Why is specifying a topic important for categorization?
Because the main category already contains 1,053 articles, specifying a topic helps organize the content more effectively for readers and editors.