請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/92072
標題: | 基於可解釋的關聯觸發詞和關係名稱在零樣本情境下進行對話關係抽取 Zero-Shot Dialogue Relation Extraction by Relating Explainable Triggers and Relation Names |
作者: | 徐澤淞 Ze-Song Hsu |
指導教授: | 陳縕儂 Yun-Nung Chen |
關鍵字: | 自然語言理解,關係抽取,對話系統,零樣本學習,持續訓練, Natural Language Understanding,Relation Extraction,Dialogue System,Zero-Shot learning,Continuous training, |
出版年 : | 2024 |
學位: | 碩士 |
摘要: | 建立對話關係抽取系統通常需要大量的標記數據,這可能在標註上耗費昂貴且耗時。為了提高可擴展性,支持多樣性並且能夠應對未知的關係抽取,本文提出了一種方法,借助觸發機制的能力並將其關聯到先前未見過的關係名稱來解決這個問題。
具體而言,我們引入了一種模型,利用觸發機制的捕捉能力,實現了零樣本對話關係抽取。我們在DialogRE數據集上進行的實驗顯示,所提出的模型對於已知和未知的關係都取得了顯著的改進。 值得注意的是,這是首次嘗試使用觸發機制的捕捉能力進行零樣本對話關係抽取,認識到所面對的數據未必總是全新的,我們的架構經過精心設計,以支持持續訓練,進而消除對模型進行完整重新訓練的必要。 我們的結果表明,這種方法對於推斷以前未見過的關係類型是有效的。總體而言,我們的研究結果突顯了該方法提升對話關係抽取系統可擴展性和實用性的潛力。 Developing dialogue relation extraction (DRE) systems often requires a large amount of labeled data, which can be costly and time-consuming to annotate. In order to improve scalability and support diverse, unseen relation extraction, this paper proposes a method for leveraging the ability to capture triggers and relate them to previously unseen relation names. Specifically, we introduce a model that enables zero-shot dialogue relation extraction by utilizing trigger-capturing capabilities. Acknowledging that the encountered data may not always be entirely new, our architecture is intricately crafted to support continual training, eliminating the necessity for a complete retraining of the model. Our experiments on a benchmark DialogRE dataset demonstrate that the proposed model achieves significant improvements for both seen and unseen relations. Notably, this is the first attempt at zero-shot dialogue relation extraction using trigger-capturing capabilities, and our results suggest that this approach is effective for inferring previously unseen relation types. Overall, our findings highlight the potential for this method to enhance the scalability and practicality of DRE systems. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/92072 |
DOI: | 10.6342/NTU202400219 |
全文授權: | 同意授權(限校園內公開) |
顯示於系所單位: | 資訊工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-112-1.pdf 授權僅限NTU校內IP使用(校園外請利用VPN校外連線服務) | 1.07 MB | Adobe PDF | 檢視/開啟 |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。