新聞立場檢索之相關研究

A Study on News Stance Retrieval

倪溥辰(Pu-Chen Ni)

指導教授 ： 鄭卜壬

國立臺灣大學/電機資訊學院/資訊網路與多媒體研究所/碩士(2019年)

https://doi.org/10.6342/NTU201901533

摘要

具爭議性議題的新聞一直是閱聽人關注與討論的焦點，例如:美國牛肉開放進口、死刑廢除、多元成家等。不論是政治、經濟、教育、兩性、能源、環保等公共議題，新聞媒體常需報導不同的立場。若能從大量的新聞文件裡，快速搜尋各種爭議性議題中具特定立場的新聞，不但有助於人們理解不同立場對這些議題的認知與價值觀，對制定決策的過程而言，也相當有參考價值。
在本篇論文中，針對新聞立場檢索這個題目做了「一條龍式」的研究，包含創建資料集、標注規則定義、比較各種可能的檢索方法。並提出一種新穎的關聯回饋機制，比起傳統的關聯回饋更具效率，而實驗證明該機制亦能提高新聞立場檢索的表現。

關鍵字

資訊檢索 ； 立場分類 ； 資訊擷取 ； 意見探勘 ； 關聯回饋

並列摘要

News with controversial topic has always been discussed by the audience. For example, US beef open for import, abolishing death penalty, and same-sex marriage, etc. Whether it is political, economic, educational, gender, energy, environmental protection and other public topics, the news media often need to report in different positions. If we can quickly search for news with specific positions from a large number of news documents, it will not only help people understand different positions, but also good for making decisions.
In this paper, we have done a study on the theme of news stance retrieval from beginning to end, including creating datasets, defining labeling rule, and comparing various possible methods. And a novel relevance feedback mechanism is proposed, which is more efficient than traditional relevance feedback, and the experiments show that the proposed mechanism can improve the performance of news stance retrieval.

並列關鍵字

information retrieval ； stance classification ； information extraction ； opinion mining ； relevance feedback

參考文獻

[1] A. Joulin, E. Grave, P. Bojanowski, M. Douze, H. J gou, and T. Mikolov. Bag of Tricks for Efficient Text Classification. arXiv preprint arXiv:1607.01759, 2016.

[2] G. Salton and C. Buckley. Term-weighting approaches in automatic text retrieval. Information Processing & Management, volume 24, issue 5, pages 513-523, 1988.

[3] G. Salton, A. Wong, and C. S. Yang. A vector space model for automatic indexing. Communications of the ACM, volume 18, issue 11, pages 613-620, 1975.

[4] G. Zarrella and A. Marsh. MITRE at semeval-2016 task 6: Transfer learning for stance detection. In Proceedings of the 10th International Workshop on Semantic Evaluation, SemEval@NAACL-HLT 2016, San Diego, CA, USA, June 16-17, 2016, pages 458–463, 2016.

[5] J. Du, R. Xu, Y. He, and L. Gui. Stance classification with target-specific neural attention. In Proceedings of the 26th International Joint Conference on Artificial Intelligence, IJCAI-17, pages 3988–3994, 2017.