The intelligent agent is a concept that has been around for approximately 25 years. The definitions of an agent fall everywhere along a continuum, from simple macros in which the user enters a few parameters to truly intelligent agents which demonstrate learning ability and artificial intelligence. Currently, most intelligent agents fall somewhere between the former description and the middle ground of the continuum. In general, though, all agents seem to have some similar characteristics. An agent is a software entity which has some degree of autonomy, carries out operations on the behalf of a user or another program, and in this process, represents or has knowledge of the user's goals and wishes. www.cs.umbc.edu/agents/introduction

An example of an agent task would be if the user wished to find the lowest price for a specific item, such as a music CD. Rather than the user having to surf the net for endless hours trying to locate the least expensive price, the agent would be free to do as he wished. Monitoring the search would not be necessary and a list of low prices with store identification would be returned.

The qualities of an agent can be broken down further into three main dimensions: agency, intelligence, and mobility. Agency describes the degree of independence which an agent exhibits. At the least, the agent must be able to operate on the Internet while the user is disconnected or not in the process of interacting with the World Wide Web. The most basic agent represents the user in some way and proceeds to complete a task with that information in mind. More complex agents can interact with other programs and data. Eventually, as the field of artificial intelligence continues to grow, this independent agent will be able to cooperate with other agents in a mutually beneficial way.

Intelligence is the amount of learned behavior and possible reasoning capacity that an agent may possess. At the most basic level, the agent may follow a set of rules that are pre-defined by the user. The agent can then apply these rules to the Internet at large. An agent who followed a model would demonstrate the next level of intelligence. A user model would not have to be as specific as a set of rules and would include a plan to retrieve the information as well. Finally, the most intelligent agents will be able to learn. They will also be able to adapt to their environment, in terms of user requests and the resources available to the agent. These agents may even make their own relationships with other agents and create new agents with higher reasoning capabilities. All of this interaction will benefit the user by providing better information, yet will be invisible to the user. Basically, their needs will be met and the agent will have taken care of the details.

Mobility is the dimension which makes agents useful in Internet applications, yet poses many privacy and security issues. An agent can be static and reside on the client computer to manage and gather information on that one machine. A more mobile agent may be composed on one computer (at the client or server end) and then sent to another to gather data. It then returns the data to the client and is sent out to gather another piece of information. This pattern repeats until the agent has exhausted the search parameters or it is stopped by the user. This arrangement has the advantage of good security.

The other option related to mobility is an agent which travels from computer to computer gathering information, but does not return to the client computer until it has collected all of the requested data. In other words, this agent accumulates and carries the data in the process of its search. While security is not as good on this form of agent, it will have the distinct advantage of allowing agents which have similar information and goals to interact. There may even be places where the agents will "meet" to share information and collaborate on their individual quests for data.

Although the issue of security is definitely a valid one which needs to be dealt with, we must remember that similar concerns were voiced ten years ago when networks were gaining in popularity. Many people though it was a poor idea to connect to a network, at the time. Today these same people probably cannot imagine doing business without being connected to their Local Area Network (LAN) or the Internet. A similar transition period of acceptance for agents is likely, yet the benefits of agents will most likely outweigh the risks. In a world of growing efficiency, agents will be an integral part of this growth.