透過您的圖書館登入
IP:216.73.216.179
  • 學位論文

通志脈絡分析系統:數位人文脈絡分析系統方法論的一個範例

Tongzhi Context Analysis Systems: A Digital Humanities-based System Design Methodology for Chinese Classics and Its Realization on Tongzhi

指導教授 : 項潔
本文將於2027/11/01開放下載。若您希望在開放下載時收到通知,可將文章加入收藏

摘要


本研究旨在探討如何運用數位人文「脈絡分析系統」的方法論來分析和呈現古籍中的多重脈絡。本研究的核心是設計並實作一個以鄭樵《通志》為分析對象的數位人文系統──通志脈絡分析系統(https://tongzhi.com.tw/)。此系統不僅提供文獻檢索和分析功能,更著重於揭示文本之間多元的關聯和脈絡,為研究者提供新的視角,以全面理解文本內容,進而開拓新的研究方向。 「脈絡分析系統(Context Analysis System)」是數位人文核心的方法論之一,它能夠盡可能地挖掘文件之間的各種關聯,呈現多重型態的脈絡,讓人文學者可以輕易的觀察到這些現象。同時,系統也為不同型態的脈絡提供其適合的分析功能,協助研究者進行後續的分析與詮釋。也就是說,脈絡分析系統務求清晰、客觀地呈現文件間各種關聯(也就是脈絡),並提供使用者一個方便觀察與分析這些脈絡的環境。 本研究所使用的文本──《通志》,由鄭樵所編纂,共計兩百卷。其中,「二十略」是全書最精華的部分。他選擇了二十個他認為重要的知識領域,將每種知識整理成一個「略」,並在「略」中對該知識領域做深入地分析。他在「二十略」中強調「會通」的核心理念,透過對不同領域知識大規模、系統性的資料彙集,在建立知識體系的過程中,掌握文本和概念間的關聯。這種方法不僅是對既有知識的整理,更是對知識的「重組」和「再發現」。該做法也讓這些文本間、概念間的關聯得以彼此延續、擴散、互動與結合,最終形成一個宏大而可觀的研究視角。 本研究詳細闡述了研究過程中資料處理的流程,包括文本的數位化、標點、後分類和詞彙標記等步驟,確保系統檢索和分析結果的可靠性和有效性。此外,系統在功能的設計上注重使用者體驗,提供直觀的操作界面和豐富的互動功能,並兼顧資料處理的效能和穩定性。系統整合了後分類、詮釋資料、詞彙標記、GIS視覺化、文本比對、AI翻譯和時間軸等多元功能,幫助研究者從不同角度觀察和分析文本,發掘其中隱含的脈絡和深層的意義。 本研究期待透過「脈絡分析系統」來重現、甚至重構《通志》的內容。希望在如實反映出鄭樵著作觀點的同時,進一步以資訊科技來發掘、呈現與分析文件之間更多的脈絡關係,藉此實現對《通志》中「會通」精神的掌握,提供研究者探索《通志》的多重視角,並期望以此拓展與深化《通志》相關研究的邊界及成果。

並列摘要


This study explores the application of Digital Humanities’ methodology of “Context Analysis Systems” in analyzing and presenting multiple contexts within Chinese classics texts. The study centers on the design and implementation of a collection of Digital Humanities systems—the Tongzhi Context Analysis Systems (https://tongzhi.com.tw/) — which uses Tongzhi as its primary subject. These systems not only provide document retrieval and analysis functions but also reveal the connections and contexts between texts, offering researchers new perspectives for a comprehensive understanding of the text, and paving the way for new research directions. The Context Analysis Systems, as a core methodology in Digital Humanities, aim to thoroughly excavate various connections between documents and present multiple forms of context, allowing Humanities scholars to easily observe these phenomena. Simultaneously, the systems provide appropriate analytical tools for different types of contexts, assisting researchers in subsequent analysis and interpretation. In essence, the Context Analysis Systems strive to clearly and objectively present various relationships (contexts) between documents while providing users with a convenient environment for observing and analyzing these contexts. The text used in this study, Tongzhi, compiled by Zheng Qiao, comprises 200 volumes. The Twenty Outlines section is considered the essence of the work. Zheng selected twenty knowledge domains he deemed significant, organizing each into an Outline and providing in-depth analysis of each domain. In the Twenty Outlines, Zheng emphasized the core concept of “Comprehensive Understanding” (會通). Through large-scale, systematic data collection across different knowledge domains, he grasped the connections between texts and concepts while establishing a knowledge system. This approach not only organizes existing knowledge but also reconstructs and rediscovers it, allowing these textual and conceptual connections to continue, expand, interact, and combine, ultimately forming a grand and impressive research perspective. This systems provide an intuitive interface and rich interactive features while balancing the reliability and effectiveness of the systems’ retrieval and analysis results. The systems integrate diverse functions such as post-classification, metadata and tag analysis, GIS visualization, text comparison, AI translation, and timelines, enabling researchers to observe and analyze texts from various perspectives and uncover implicit contexts and deeper meanings. This study aims to recreate and potentially reconstruct the content of Tongzhi through the Context Analysis Systems. While accurately reflecting Zheng Qiao's research perspectives, it further seeks to discover, present, and analyze more contextual relationships between documents. By leveraging information technology, the study expands the conceptual boundaries of "Comprehensive Understanding" and "Context Analysis" in the field of Digital Humanities.

參考文獻


王汎森(2014)。數位人文學之可能性及限制──一個歷史學者的觀察。數位人文研究與技藝,頁25-35。
王泰升(2011)。數位化歷史資料庫與歷史研究──以明清檔案、淡新檔案、日治法院檔案等資料庫為例 。從保存到創造:開啟數位人文研究,頁31-50。
王浩宇(2024)。史料文字比對方法與系統之建置研究──以《通志》為例〔未出版之碩士論文〕。國立臺灣大學電機資訊學院網路與多媒體研究所。
王曉真(2022)。北魏神龜二年兵變與「武人入選」探微。魏晉南北朝隋唐史資料,第四十六輯,頁183-209。
毛漢光(1986)。北魏東魏北齊之核心集團與核心區。中央研究院歷史語言研究所集刊,第57本第2分冊,頁278-280。

延伸閱讀