摘要：We propose a framework for answering the question of whether statements made by an agent can be believed, in light of observations made over time. The basic components of the framework are a formalism for reasoning about actions, changes, and observations and a formalism for default reasoning. The framework is suitable for concrete implementation, e.g., using answer set programming for asserting the truthfulness of statements made by agents, starting from observations, knowledge about the actions of the agents, and a theory about the "normal" behavior of agents.
