Model for Dialogue between Informational Agents

Abstract. A theoretical model for agents interaction to study semantic heterogeneity is proposed. The context of this study is informational agents with knowledge bases that are heterogeneously interacting in a dialogue. Our problem is then: how does the interaction influence the evolution of the semantic heterogeneity between the agents, and how does this heterogeneity influence the interaction. To study semantic heterogeneity, we define agents that are not embedded in a common task program, they have no common goal and no common domain vocabulary. Our goal is then double:propose a model of dialogue between semantically heterogeneous informational agents, and use it to study semantical heterogeneity (HS) per se in a dynamical context in order to propose scenarios for solving this HS.

• Type: Conference
• Year: 2003
• Where: EPIA
• Authors: Erika Valencia, Jean-Paul Sansonnet
• Keywords: EPIA 2003, Heterogeneity Influence, Informational Agents, Semantic Heterogeneity