A Retrieval-Augmented Generation with Dual-Similarity Monitoring for Nuclear Energy Knowledge Q&A

We present a Retrieval-Augmented Generation (RAG)-based question-answering system for nuclear energy science communication, characterizing retrieval quality in generated responses. The system introduces a dual-similarity analysis that jointly measures (i) question-to-context (Q→C) and (ii) answer-to...

Бүрэн тодорхойлолт

Номзүйн дэлгэрэнгүй

Үндсэн зохиолчид:	Cheng-Hsing Chiang, Kun-Chou Lee
Формат:	Artigo
Хэл сонгох:	Inglês
Хэвлэсэн:	MDPI AG 2026-03-01
Цуврал:	Applied Sciences
Нөхцлүүд:	retrieval-augmented generation FAISS LangChain semantic similarity semantic alignment nuclear energy education
Онлайн хандалт:	https://www.mdpi.com/2076-3417/16/7/3182
Шошгууд:	Шошго нэмэх Шошго байхгүй, Энэхүү баримтыг шошголох эхний хүн болох!