Computational Dialectics Based on Specialization and Generalization --- A New Reasoning Method for Conflict Resolution

  • Authors:
  • Hiroyuki Kido;Masahito Kurihara

  • Affiliations:
  • Graduate School of Information and Science, Hokkaido University,;Graduate School of Information and Science, Hokkaido University,

  • Venue:
  • New Frontiers in Artificial Intelligence
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

The purpose of this paper is two-fold. Firstly, to formalize one aspect of dialectical thought, i.e., our way of thinking about conflict resolution. Secondly, to show that this way of thinking allows agents to resolve a conflict by argumentation. To this end, we propose a dialectical reasoning method by means of specialization and generalization defined by a logical implication. This method has three features. First, it does not limit its premises to logical contradictions in accordance with philosophical knowledge that an antithesis is not adequately expressed as logical negation of a thesis. Second, it embraces our actual and familiar thoughts exemplified in this paper. Third, it has the ability to draw conclusions that are not just logical deductions from its premises. Further, by applying it to argumentation, we show that it allows agents to resolve a conflict by drawing an alternative solution not deduced from any consistent subset of the union of all agents' knowledge base. In other words, it allows agents to develop argumentation dialogically in terms of producing an alternative solution that is not obtained at the beginning of argumentation.