Knowledge Graph Construction for Cityscape Evaluation Theories Empowered by Artificial Intelligence
叶宇 1武静芬 2李翔宁3
扫码查看
点击上方二维码区域，可以放大扫码查看
作者信息
1. 香港大学;同济大学建筑与城市规划学院;同济大学高密度人居环境与生态节能教育部重点实验室
2. 同济大学建筑与城市规划学院
3. 同济大学;同济大学建筑与城市规划学院
折叠
摘要
由于城市风貌构成及其评价理论系统认知的缺乏,中国快速城镇化进程中城市"千城一面"的现象日益明显.本研究结合检索增强生成与提示工程的技术优势,探索了人工智能赋能下的城市风貌评价理论知识图谱的构建路径,进而实现城市风貌评价理论的知识映射与深化认知.一方面为城市风貌管控与设计决策提供有效支持,另一方面为"AI for Architecture"新范式的建构提供理论层面的赋能.
Abstract
The lack of a systematic understanding of cityscape and its evaluation theories has contributed to the growing phenomenon of"uniformity across cities'"in China during rapid urbanisation.This study leverages the strengths of retrieval-augmented generation and prompt engineering to explore the construction of knowledge graphs for cityscape evaluation theories based on artificial intelligence,which enables the deepening of knowledge in the field of cityscape evaluation.In the long term,this study not only offers crucial support for cityscape management and design decision-making but also provides a solid theoretical foundation for advancing the new"AI for Architecture"paradigm.
关键词
人工智能/城市风貌评价/知识图谱/大语言模型/检索增强生成/提示工程
Key words
artificial intelligence/cityscape evaluation/knowledge graph/large language model/retrieval-augmented generation/prompt engineering
维普
万方数据