Collective assessment of arguments in an online intelligent argumentation system for collaborative decision support

Abstract

We developed an intelligent argumentation system for collaborative decision support which allows stakeholders to exchange arguments, captures their rationale, and assesses aggregate thoughts on solution alternatives [1, 21]. In this article we propose and evaluate a method for computing aggregate thoughts of arguments on individual arguments, which represents individual thoughts, in an argumentation tree and further classifying arguments based on relationships between its aggregate thoughts and individual thoughts. A novel approach is developed to derive collective determination of an argument based on total support and attack that an argument receives in the argumentation tree. The collective determination value and the strength of an argument are used to analyze the relationship between aggregate thought and individual thought of arguments and cluster the arguments. The decision maker will be able to review clusters of arguments with opposing collective thoughts and supporting collective thoughts versus their individual thoughts. We have carried out an experiment to evaluate the proposed method, and the experimental results show that the proposed method is effective.