Software agents as boundary objects

Abstract

Despite the wide use of agent-based applications in different areas of human activity, there hasn’t been paid much attention to understand how these applications are possible, taking into account that they are build by people coming from such conceptually distant fields of study as, for example, law, artificial intelligence, and software engineering. This paper aims to fill in this gap addressing the different approaches to software agents—understood as building blocks of agent-based applications—adopted in each of these fields of study and suggesting that the way to understand how do these fields manage to work together in building a single agent-based application resides in seeing these agents as boundary objects.