18-01-2011, 12:56 PM
Abstract
"Agent" and "agent software" have become popular words in computer software. The reason why this area gains the popularity is because it is based upon AI (Artificial Intelligence) but works only in its specific field, like a narrowly focused AI program. Agent software has been developed for many different uses in a variety of areas because of its extraordinary ability of adapting to the specific field of interest. This project first presents an overview of the software agent, which starts from the introduction of the definition of the software agent, its terminology, its formal frameworks and its other basics. Afterwards, the concepts and specifications of 'smart' agent framework are brought into context to describe how the software agent works behind the scene. This includes action, operation, autonomous behavior and communication of agents. The way that a software agent generates its goals of achievement and to evaluation of its progress is also very important in keeping agent working in the right path. Next, we discuss the new generation of intelligent software agents to take a brief look at how an intelligent software agent can complete its tasks and what it can do in the present time. The Growing Book project is a very typical software agent application that helps distance learning students obtain the correct course material. With the correct input of a username and password, the students will have all the class material available for them in a short time on their local computer. This saves a lot of work for students and it also greatly decreases the chance that students may download the wrong material. On the other hand, this agent protects the class material file server from access by outsiders. A scenario of the Growing Book software agent is included
