方法对比

	语篇分析	论证挖掘	情感分析
领域	文本挖掘	文本挖掘	文本挖掘
方法族	Process / pipeline	Process / pipeline	Process / pipeline
起源年份	1988 (RST); 2008 (PDTB 2.0)	2016	—
提出者	Mann & Thompson (RST); Prasad et al. (PDTB)	Lippi & Torroni (state-of-the-art survey)	—
类型	NLP discourse-structure analysis task	NLP information-extraction task	NLP text-classification task
开创性文献	Mann, W. C. & Thompson, S. A. (1988). Rhetorical Structure Theory: Toward a functional theory of text organization. Text, 8(3), 243-281.	Lippi, M. & Torroni, P. (2016). Argumentation Mining: State of the Art and Emerging Trends. ACM Transactions on Internet Technology, 16(2), Article 10, 1-25.	Pang, B. & Lee, L. (2008). Opinion Mining and Sentiment Analysis. Foundations and Trends in Information Retrieval, 2(1-2), 1-135.
别名	rhetorical structure analysis, RST parsing, PDTB parsing, Söylem Ayrıştırma (Discourse Parsing)	argumentation mining, argument extraction, Argüman Madenciliği	opinion mining, polarity detection, duygu analizi
相关	3	4	3
摘要	Discourse parsing is a natural-language-processing task that models the rhetorical relations between sentences and paragraphs of a text — relations such as cause, contrast, and elaboration — and represents them as a tree structure. It works within established frameworks, principally Rhetorical Structure Theory (RST), introduced by Mann and Thompson in 1988, and the Penn Discourse TreeBank (PDTB), released by Prasad and colleagues in 2008.	Argument mining is a natural-language-processing task that automatically detects claims, premises and the argumentative structures that link them within text. Consolidated as a field by Lippi and Torroni's 2016 state-of-the-art survey, it is applied to scientific writing, legal documents and debate analysis to turn free-form argumentation into structured, analysable units.	Sentiment analysis, also called opinion mining, is a natural-language-processing task that detects the emotional tone of text — typically classifying it as positive, negative, or neutral. It turns unstructured opinion text into structured, quantifiable polarity signals using one of three families of approaches: sentiment lexicons, trained machine-learning classifiers, or pretrained transformer models.