CORK: A COnversational agent framewoRK exploiting both rational and emotional intelligence

This paper proposes CORK, a modular framework to facilitate and accelerate the realization and the maintenance of intelligent Conversational Agents with both rational and emotional capabilities. A smart CA can be integrated in any digital device and aims at interpreting the natural language inputs by the user and at responding to them consistently with the semantics, the context, the user’s perceived emotional state and her/his profile. CORK’s strength is its ability to split the content of the speech from its pure conversational scheme and rules: it permits to the developers to detect during design some general communicative patterns to be filled at run-time with the right content depending on the context. At first, this can be tiresome and time-consuming, but it permits then to achieve high scalability and low-cost modification and maintainability. This framework is potentially valid in general and for now it has been effectively applied to develop conversational tools for people with neurodevelopmental disorders.