A survey of argumentation structures for intelligent decision support

In practical reasoning, knowledge is often used to argue, in various ways, for or against an assertion. This perspective on knowledge has led a number of researchers to apply insights from argumentation theories to modelling reasoning. However, argumentation theories have been applied to the development of intelligent systems based in a variety of ways. For example, the most commonly used structure for representing knowledge in this way is the Toulmin Argument Structure (TAS) yet many variations to Toulmin’s original structure have been developed. In this paper, we survey different argumentation approaches including our own TAS variation in a family law system, Split Up, in order to conclude that there is unlikely to be one universally optimal argumentation framework but that each application requires a variation to suit the task.