Agenti artificiali e agenti intelligenti: paradigmi, applicazioni e prospettive

The concept of agency in computer science is related to computational entities, called artificial agents, which are able to interact with each other or with human beings. Intelligent agents are specific agents that perform tasks and activities requiring reasoning, thus showing intelligent behaviours. The objective of this paper is providing an overview of this research field, considering its historical development, the proposed approaches and the main applications. The paper discusses a possible definition of agency in the computer science research, presenting the two main approaches to the development of agents in Artificial Intelligence, i.e., goal-driven deliberative agents, and reactive agents. Within these two main research threads, various architectures are discussed, including hybrid ones. Multi-agent systems are also considered, with their emphasis on communication among interacting agents. The paper ends up with a brief presentation of some possible present and future perspectives, with particular attention to the possibilities offered by the Web.