Retrieval in seconds
Vector retrieval across all three corpora; a glossary hit constrains the output right away.
Three corpora + AI live Q&A and ten-language translation - minute-level language support for the disaster scene.
Need a term right now? Open the glossary
Raise the request - ask a question on the Q&A page or paste text into machine translation.
The query is embedded and matched by cosine similarity across the news, parallel corpus and glossary; the hits become the authoritative context, and matched glossary terms act as hard constraints.
Vector retrieval across all three corpora; a glossary hit constrains the output right away.
Q&A and translation stream token by token, so the answer starts before the request ends.
Every answer and translation lists the glossary, corpus and news items behind it.
Chinese, English and eight ASEAN languages, all in one workbench.
Browse, search and edit the 10 INSARAG news records (bilingual JSON).
Aligned EN/ZH passages from INSARAG Guidelines Volume I: Policy, serving as translation memory.
Authoritative INSARAG terminology with abbreviations, categories and notes.
Ask in natural language; answers are grounded in all three corpora with traceable sources.
Translate among 10 ASEAN-focused languages, grounded by term base, corpus and news.
On site you look a term up, read it correctly and turn an instruction into another language right away.
Describe the need in plain language and get an answer grounded in authoritative corpora, with its sources attached.
Prepare lessons and exercises against the aligned EN/ZH corpus and past news cases.