查詢結果分析
來源資料
相關文獻
- 生成式AI於古典詩敘事轉圖像之語境構築與文化設計應用研究
- 博物館與慶典: 人類學文化再現的類型與政治
- 情景法的理論與應用--以中學詩歌課文為例
- Beyond Assimilationism and Separatism: A Multicultural Study on M. H. Kingston's Tripmaster Monkey: His Fake Book
- 談虛實(情景)法在高中詩詞課文中的運用
- 族群展示的反思:後威權臺灣的觀察
- 當戰場變成情場:青少年電影《夏天協奏曲》的文化再現
- 苗栗縣巴宰族群文化再現與復振之研究
- 從當代「文化再現」的詮釋觀點談「宋家皇朝」的「去強人」情結
- 鄉土食和山水亭:戰爭期間「臺灣料理」的發展(1937~1945)
頁籤選單縮合
| 題 名 | 生成式AI於古典詩敘事轉圖像之語境構築與文化設計應用研究=Contextual Construction and Cultural Design Applications of Generative AI in the Visual Transformation of Classical Poetic Narratives |
|---|---|
| 作 者 | 莊育鯉; | 書刊名 | 海洋文化學刊 |
| 卷 期 | 39 2025.12[民114.12] |
| 頁 次 | 頁79-114 |
| 分類號 | 312.83 |
| 關鍵詞 | 生成式AI; 情景法; 詩意設計; 文化再現; Generative AI; Prompt engineering; Context engineering; Scenario method; Poetic design; Cultural representation; |
| 語 文 | 中文(Chinese) |
| 中文摘要 | 本研究探討生成式 AI(Generative AI)在設計領域中的語法轉向,提出由 「Prompt Engineering」進化至「Context Engineering」的理論架構,並以清代李逢 時詩作〈社寮漁火〉為文化原型,實踐 AI 輔助的詩意包裝設計。研究旨在釐清語 法層的 AI 生成設計侷限,進而建構具文化語境與情感深度的設計生成模式。在理 論層面,本研究整合修辭學「章法結構」與「情景法」的概念,建立 AI 設計中的 三層架構模型:語法層(Syntactic Layer)、語境層(Contextual Layer)與設計層 (Design Layer)。研究透過文獻回顧、語法與語境生成比較實驗,以及三位文學與 美學專家之質性訪談,歸納 AI 在詩意再現、地方文化詮釋與設計敘事建構中的轉 譯機制與文化價值。結果顯示,「Context Engineering」能有效強化 AI 生成圖像的 章法層次感與敘事深度,使設計成果呈現出兼具詩學意境、文化符號與視覺美感 的整體結構。在實務應用上,本研究以基隆和平島的地方文化為設計場域,透過 AI 生成重構〈社寮漁火〉詩中「漁燈風亂搖」的意象,結合地方產業「石蓴」之 視覺符號,發展出兼具文化敘事與商品功能的詩意包裝設計。專家分析指出,本 研究成功展現 AI 生成設計的人文深度與教育價值,亦揭示 AI 在文化創意產業中 作為「語境建構者(context builder)」的新角色。研究證實生成式 AI 不僅可作為 視覺生成工具,更能成為文化詮釋與情感傳達的媒介。「Context Engineering」的設 計轉向不僅擴展了 AI 生成的詩學維度,也為 AI 設計教育與文化永續發展提供新 的理論基礎與實踐方向。 |
| 英文摘要 | This study explores the syntactic shift in generative design practices enabled by Generative Artificial Intelligence (Generative AI), proposing a theoretical framework that advances from Prompt Engineering to Context Engineering. Drawing on Li Feng-shi’s Qing dynasty poem “Fishing Lights at Sheliao” as the cultural prototype , the research employs AI-assisted poetic packaging design to investigate how AI can integrate cultural context and emotional depth into the generative design process. Theoretically, this study integrates the rhetorical principles of composition and the scenario method to establish a three-layer framework for AI design: the Syntactic Layer, the Contextual Layer, and the Design Layer. Through a literature review, comparative experiments on syntactic versus contextual generation, and qualitative interviews with three experts in literature and aesthetics, the study identifies the mechanisms and cultural values underlying AI’s translation of poetic imagery, local cultural interpretation, and design-based storytelling. The findings indicate that Context Engineering effectively enhances the compositional hierarchy and narrative depth of AI-generated imagery, yielding design outcomes that embody poetic aesthetics, cultural symbolism, and visual coherence. In practical application, this study focuses on the local culture of Keelung’s Heping Island as the design context. By employing AI to reinterpret the imagery of “the fishing lights swaying in the wind” from “Fishing Lights at Sheliao,” and integrating visual elements inspired by the local sea lettuce (Ulva) industry, the research develops a poetic packaging design that unites cultural narrative with commercial functionality. Expert analyses confirm that the project demonstrates the humanistic depth and educational significance of AI-generated design, revealing AI’s emerging role as a “context builder” within the cultural and creative industries. In conclusion, this study affirms that Generative AI is not merely a tool for visual creation, but also a medium for cultural interpretation and emotional communication. The proposed transition toward Context Engineering broadens the poetic and cultural dimensions of AI-generated design, establishing a new theoretical foundation and practical orientation for AI design education and sustainable cultural development. |
本系統中英文摘要資訊取自各篇刊載內容。