A Model of Culture for Cognitive Agents

A Mas For Managing Cultural Conflicts

The work described in this paper is part of a project aimed at supporting collaboration in multicultural teams. In particular, we intend to reduce the impacts of cultural misunderstandings by means of the construction of cultural profiles for users and the analysis of interpersonal interactions. The overall solution proposed for this problem is a multi-agent system (MAS), based on the OMAS architecture (Barthès, 2011), which provides advanced cognitive agent functionalities. The OMAS platform provides three kinds of built in agents: personal assistants, service agents, and transfer agents. Personal assistants are in charge of interfacing users, service agents provide diverse services depending on the application, and transfer agents are in charge of network communications and interfacing external systems like other multi-agent systems or web services. In the particular case of our project, which highly depends on user interactions, OMAS built-in functionalities provided by personal assistants are especially useful. Such functionalities even allow users to communicate with agents in the system through a natural language, vocal interface.