A Formal System for Lies Based on Speech Acts in Multi-Agent Systems

In the field of multi-agent system the research of speech acts is both interesting and important. In the classical theory of speech acts, the sincerity principle is regarded as a compulsory condition. However, the sincerity principle can not be guaranteed in many multi-agent systems because of the existence of lies in the speech interactions among agents. This paper discusses this issue from the perspective of cognition processes of agents, gives the definition of lies based on speech acts and formalizes these contents with LOBA (logic of believable agents). LOBA has extended the work of BDI logic, LORA logic and KARO logic, and introduced so-called cognitive actions which only occur in agents´ minds to describe the dynamic cognition processes of believable agents