Abstract
Cognitive agent communication is a research field in full development. We propose here an extension and a partial implementation of the STROBE model, which regards the agents as Scheme interpreters. These agents are able to interpret messages in a dedicated environment including an interpreter that learns from the current conversation. These interpreters evolve dynamically, progressively with the conversations, and thus represent evolving meta-level agent's knowledge. We illustrate this theoretical model by a “teacher-student” dialogue experimentation, where an agent learns a new performative at the completion of the conversation.This report is organized as follows. First, we survey the two research domains inspiring our experiment: Scheme evaluation and agent communication. Then we present our model and illustrate it by an experimentation. Finally, we show how this model is effective for domains such as the Web, Grid Computing and e-commerce dialogue (with a constraint point of view). This last idea is particularly developed to show how to enable the dynamic specification of a problem by designing communicating agents. The implementation details are also available.