透過您的圖書館登入
IP:44.220.131.93
  • 學位論文

自動鏈結資料產生器之發展與數位人文教育應用研究

Development of an Automatic Linked Data Generator and Its Application in Digital Humanities Education

指導教授 : 陳志銘
本文將於2027/02/16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


本研究旨在開發支援數位人文探究之「自動鏈結資料產生器」,以輔助數位人文學習者在進行大量文本閱讀時,能藉由文本關聯推薦快速掌握及解讀文本內容,以利於梳理出相關人、事、物之間的關聯脈絡。同時,藉由相關文章之文章摘要提供遠讀和細讀相互鏈結的功能,以利於數位人文學習者能更有效率地在相關聯的文章之間進行探索。為了驗證此一工具對於支援數位人文探究之效益,本研究邀請具中文、歷史、哲學等相關背景共16位學生為實驗對象,以實驗研究法之對抗平衡設計比較實驗對象依序使用有/無「自動鏈結資料產生器」之「羅家倫先生文存數位人文平台」輔助進行不同面向文本探索,在文本探索摘要短文分數及科技接受度是否具有顯著的差異,並以半結構式深度訪談瞭解實驗對象對「自動鏈結資料產生器」的看法與感受,最後透過滯後序列分析探討實驗對象使用「具自動鏈結資料產生器」之「羅家倫先生文存數位人文平台」輔助進行不同面向文本探索的有效行為轉移模式。 實驗結果發現,實驗對象採用有/無「自動鏈結資料產生器」之「羅家倫先生文存數位人文平台」輔以進行文本探索,其探索文本成效依據探索之文本主題不同而有不同的顯著差異。「具自動鏈結資料產生器」之「羅家倫先生文存數位人文平台」較適合於探索歷史類主題文本,而「無自動鏈結資料產生器」之「羅家倫先生文存數位人文平台」較適合於探索教育類主題文本。在科技接受度分析部分,採用「具自動鏈結資料產生器」之「羅家倫先生文存數位人文平台」在認知有用性上顯著高於「無自動鏈結資料產生器」之「羅家倫先生文存數位人文平台」,表示實驗對象認為「自動鏈結資料產生器」更能輔助其在文本探索中,有效率地掌握相關聯的文章與可探討的主題。在行為分析得知,交互使用「自動鏈結資料產生器」與傳統關鍵字檢索及後分類功能,為比較有效的系統操作行為模式。此外,透過訪談結果得知,大多數實驗對象認為「自動鏈結資料產生器」所產生的相關文章之摘要,提供了遠讀和細讀相互鏈結的功能,可以幫助他們更有效率地在相關聯的文章之間進行探索。在未來研究方向上,可提供文章與文章之間人物的社會網絡關係圖,以藉由人物關係進行文本探索,並發展人機互動的合作模式,提升文本實體標註的準確性,以及嘗試以主題分析(topic analysis)(Pan & Li, 2010)的方式來建立文章之間的關聯,以提供文本探索上更多其他不同面向關係文章的連結性。

並列摘要


This research aims to develop an “Automatic Linked Data Generator” on the Digital Humanities Research Platform for Mr. Lo Chia-lun’s Writings in order to assist learners in quickly grasping and interpreting text contents through the texts recommended by the generator for supporting digital humanities education. It can assist learners to explore the texts relating to a certain topic after viewing the linked data relation map provided by “Automatic Linked Data Generator”, so that learners can more efficiently explore interested topics from a huge amount of texts on Digital Humanities Research Platform for Mr. Lo Chia-lun’s Writings. To verify the effectiveness of “Automatic Linked Data Generator” in supporting the exploration of digital texts for a target topic, a total of 16 students were recruited as the research participants. The counterbalanced design in quasi-experimental research was applied in this study to compare whether the text exploration effectiveness and technology acceptance of the research subjects with and without the support of “Automatic Linked Data Generator” on Digital Humanities Research Platform for Mr. Lo Chia-lun’s Writings were significant differences. Additionally, lag sequential analysis was also used to analyze learners’ operation behaviors on the Digital Humanities Research Platform for Mr. Lo Chia-lun’s Writings with “Automatic Linked Data Generator.” A semi-structured in-depth interview was also applied to understand learners’ opinions and perception of using the Digital Humanities Research Platform for Mr. Lo Chia-lun’s Writings with “Automatic Linked Data Generator” to interpret texts through reading. The results of the experiment reveal that the text exploration effectiveness of the research subjects with and without the support of “Automatic Linked Data Generator” on Digital Humanities Research Platform for Mr. Lo Chia-lun's Writings has statistically significant difference and the text exploration effectiveness of the research subjects depends on the text topic explored. Particularly, the analytical results show that the Digital Humanities Research Platform for Mr. Lo Chia-lun's Writings with the “Automatic Linked Data Generator” is more suitable for exploring historical texts, while the Mr. Lo Chia-lun's Works Digital Humanities Research Platform without the “Automatic Linked Data Generator” is more suitable for exploring educational texts. The perceived usefulness in the technology acceptance of the Digital Humanities Research Platform for Mr. Lo Chia-lun's Writings with “Automatic Linked Data Generator” is significantly higher than the Digital Humanities Research Platform for Mr. Lo Chia-lun's Writings without “Automatic Linked Data Generator.” The results indicated that the “Automatic Linked Data Generator” can support learners to efficiently grasp related texts and explore topics during text exploration. In the behavioral analysis, it was found that the interactive use of the “Automatic Linked Data Generator” and the traditional keyword retrieval and post-categorization functions were more effective modes of system operation. In addition, the interview results showed that most of the research participants agreed with that the texts recommended by the “Automatic Linked Data Generator” could help them explore the target topic more efficiently. In the future, the social network relationship graph of characters between texts can be provided to assist learners in exploring the target topic through character relationship and to develop a cooperative model of human-computer interaction to enhance the accuracy of textual entity annotation. It is also possible to try to establish relationship between texts by topic analysis (Pan & Li, 2010), in order to provide more links to other texts with different relationship in text exploration.

參考文獻


Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. doi:10.2307/249008
參考文獻
一、中文文獻
中國哲學書電子化計劃(2021)。中國哲學書電子化計劃。上網日期:110年4月18日,檢自:https://ctext.org/zh
王聿均(1989)。羅志希先生對史學與文學的貢獻。載於羅家倫先生文存第十二冊(頁901-933)。臺北縣:國史館。

延伸閱讀