The ever-increasing volume of persuasive text to be analysed has driven efforts to automate the identification and reconstruction of the argumentative content contained within. Computational techniques such as opinion mining and sentiment analysis make it possible to identify the views expressed in a text, yet ignore the reasoning underpinning them. Argument mining addresses this gap by automatically reconstructing the reasons for and against claims in persuasive texts. Automatically identifying this inferential structure and its associated premises and conclusions, does not just convey what views are being expressed, but also why. The authors give an overview of the field of argument mining as the computer-based analysis of the logogical means of persuasion: the reasons advanced in justifying (or refuting) a disputed standpoint. After a short introduction to the methods of argument mining, five aspects of it are discussed in more depth: the characteristic features of persuasive language; rhetorical figures of speech; statement types used in persuasive text; argument schemes as common patterns of reasoning; and persuasive dialogues. The authors also discuss the future of argument mining, considering new more accurate methods of extraction, and potential future applications of this technology.
|Title of host publication||The Routledge Handbook of Language and Persuasion|
|Editors||Jeanne Fahnestock, Randy Allen Harris|
|Place of Publication||London|
|Number of pages||16|
|Publication status||Published - Sept 2022|