A belief representation for understanding deception
Greg B Taylor, Stephen B. Whitehill · International Joint Conference on Artificial Intelligence · 1981
Identifying deception in stories requires an understanding of the beliefs of the characters. A model must include both beliefs about facts and beliefs about other characters. This paper presents a method for representing such belief structures. Arbitrary levels of embedded beliefs are represented by cvclic structures with Shared common beliefs. With these structures we show how different instances of deception can be recognised with a single deception template. We will illustrate these concepts by applying them to several example stories.