Argumentation in the intelligence domain

Aiming to explain past events or predict future events, intelligence analysts reason about collections of often unreliable, ambiguous, and incomplete evidence to support or reject alternative hypotheses. Argumentation plays an important role in both the collaborative reasoning process within teams of analysts, and the final documents in which the outcomes of the analysis are reported. In this contribution, we explore the commonalities between standpoints and arguments on the one hand and hypotheses and evidence on the other. We will discuss the notion of analytical rigour of intelligence assessments and its relation to argument quality. With the objective of computationally representing structures of hypothetical reasoning with evidence, we discuss a machine-readable format for doing so, and we reflect on the consequences of this approach. The chosen file format serves to facilitate the development of software tools for supporting analysts in their often complex intelligence assessments in this high-stakes argumentative domain.