An Ontology-Driven MCP Agent Framework for Ensuring Integrity in Architectural Structural Design

The rapid proliferation of Large Language Model (LLM)-based autonomous agents has demonstrated strong potential for automating knowledge-intensive tasks. However, their probabilistic reasoning often leads to hallucinations and weak domain knowledge integrity, limiting their applicability in high-sta...

Полное описание

Библиографические подробности

Главные авторы:	Inkwan Paik, Dongkwan Kim, Junhyuk Lim, Youngsook Roh, Seunguk Na
Формат:	Artigo
Язык:	Inglês
Опубликовано:	IEEE 2026-01-01
Серии:	IEEE Access
Предметы:	Model context protocol (MCP) ontology knowledge integrity construction engineering architectural structural
Online-ссылка:	https://ieeexplore.ieee.org/document/11475708/
Метки:	Добавить метку Нет меток, Требуется 1-ая метка записи!