Persuasion-enhanced computational argumentative reasoning through argumentation-based persuasive frameworks

Ramon Ruiz-Dolz (Lead / Corresponding author), Joaquin Taverner, Stella M. Heras Barberá, Ana García-Fornes

Research output: Contribution to journalArticlepeer-review

4 Downloads (Pure)

Abstract

One of the greatest challenges of computational argumentation research consists of creating persuasive strategies that can effectively influence the behaviour of a human user. From the human perspective, argumentation represents one of the most effective ways to reason and to persuade other parties. Furthermore, it is very common that humans adapt their discourse depending on the audience in order to be more persuasive. Thus, it is of utmost importance to take into account user modelling features for personalising the interactions with human users. Through computational argumentation, we can not only devise the optimal solution, but also provide the rationale for it. However, synergies between computational argumentative reasoning and computational persuasion have not been researched in depth. In this paper, we propose a new formal framework aimed at improving the persuasiveness of arguments resulting from the computational argumentative reasoning process. For that purpose, our approach relies on an underlying abstract argumentation framework to implement this reasoning and extends it with persuasive features. Thus, we combine a set of user modelling and linguistic features through the use of a persuasive function in order to instantiate abstract arguments following a user-specific persuasive policy. From the results observed in our experiments, we can conclude that the framework proposed in this work improves the persuasiveness of argument-based computational systems. Furthermore, we have also been able to determine that human users place a high level of trust in decision support systems when they are persuaded using arguments and when the reasons behind the suggestion to modify their behaviour are provided.
Original languageEnglish
Pages (from-to)229-258
Number of pages30
JournalUser Modeling and User-Adapted Interaction
Volume34
Issue number1
Early online date19 Jun 2023
DOIs
Publication statusPublished - Mar 2024

Keywords

  • Computational argumentation
  • Computational persuasion
  • Human–computer interaction
  • User modelling

ASJC Scopus subject areas

  • Education
  • Human-Computer Interaction
  • Computer Science Applications

Fingerprint

Dive into the research topics of 'Persuasion-enhanced computational argumentative reasoning through argumentation-based persuasive frameworks'. Together they form a unique fingerprint.

Cite this