Intelligent Agents : An Overview

An intelligent agent (IA) is a computer program that helps a user with routine computer tasks. It performs a specific task based on predetermined rules and knowledge stored in its knowledge base. IAs are a powerful tool for overcoming the most critical limitation of the Internet information overflow and making electronic commerce a viable organizational tool.

DEFINITIONS

The term agent is derived from the concept of agency, referring to employing someone to act on your behalf. A human agent represents a person and interacts with others to accomplish a predefined task.

The concept of agents goes surprisingly far back. More than 50 years ago, Vannevar Bush envisioned a machine called a memex. In his mind, he could see the memex assisting humans through huge fields of data and information. In the 1950s, John McCarthy developed Advice Taker, a software robot that would navigate the networks of information that were expected to develop in time. Advice Taker's similarity to today's agents is amazing. Given a task by a human user, the robot takes the necessary steps or asks for advice from the user when it gets stuck.

The futuristic prototypes of intelligent personal agents, such as Apple's Phil and Microsoft's Bob, perform complicated tasks for their users following the same functions laid out by McCarthy in Advice Taker. The modern approach to intelligent agents moved to mobile and multiple agents in the mid 1980s under research topics like distributed artificial intelligence (Bond and Gasser, 1988) and agency theory.

During the development process, several names have been used to describe intelligent agents, including software agents, wizards, software demons, knowbots, and softbots (intelligent software robots). These terms sometimes refer to different types of agents or agents with different intelligence levels.

Demons were a popular term for agents in the early stage of development. A demon is a small computer program that runs in the background and takes actions to alert the user of a certain situation when a prespecified condition is met. An example is the X windows program xbiff. This program continually monitors a user's incoming e-mail, and indicates via an icon whether there are any unread messages. Virus-detection agents and incoming-e-mail agents are similar examples. Recently, the term bot has become a common substitute for the term agent. Bot is an abbreviation for robot. Bots are given specific prefixes indicating their use. Typical bots are chatterbots, docbots, hotbots, jobbots, knowbots, mailbots, musicbots, shopbots, spiderbots, spambots, and sexbots (of course it had to be).

There are several definitions of what an intelligent agent is. Each definition explicates the definer's perspective. Here are some examples:

• Intelligent agents are software entities with some degree of independence or autonomy that carry out some set of operations on behalf of a user or another program and in so doing employ some knowledge or representation of the user's goals or desires.
• Autonomous agents are computational systems that inhabit some complex dynamic environment, sense and act autonomously in this environment, and by doing so realize a set of goals or tasks for which they are designed.
• Intelligent agents continuously perform three functions: perception of dynamic conditions in the environment, action to affect conditions in the environment, and reasoning to interpret perceptions, solve problems, draw inferences, and determine actions.
• A software implementation of a task in a specified domain, on behalf or in lieu of an individual or other agent. The implementation will contain homeostatic goal(s), persistence, and reactivity, to the degree that the implementation will persist long enough to carry out the goal(s), and will reach sufficiently within its domain to allow the goal(s) to be met and to know that fact.
• An agent is a computer system that is situated in some environment and is capable of autonomous action in this environment in order to meet its design objectives.

INTELLIGENCE LEVELS

Intelligence is a key feature related to defining intelligent agents because it differentiates them from ordinary agents. Wooldridge (2002) suggested that intelligence in this sense possesses the following features:

• Reactivity. Intelligent agents are able to perceive their environment and respond in a timely fashion to changes that occur in it in order to satisfy their design objectives.
• Proactiveness. Intelligent agents are able to exhibit goal-directed behavior by taking the initiative in order to satisfy their design objectives.
• Social ability. Intelligent agents are capable of interacting with other agents (and possibly humans) in order to satisfy their design objectives.

There are four different levels of agent intelligence, as follows:

Level 0 (the lowest). These agents retrieve documents for a user under straight orders. Popular Web browsers fall into this category. The user must specify the URLs where the documents are. These agents help in navigating the Web.

Level 1. These agents provide a user-initiated searching facility for finding relevant Web pages. Internet search agents such as Google, Alta Vista, and Infoseek are examples. Information about pages, titles, and word frequency is stored and indexed. When the user provides key words, the search engine matches them against the indexed information. These agents are referred to as search engines.

Level 2. These agents maintain user profiles. Then they monitor Internet information and notify the users whenever relevant information is found. An example of such agents is WebWatcher, the tour-guide agent for the Web developed at Carnegie Mellon University. Agents at this level are frequently referred to as semi-intelligent or software agents.

Level 3. Agents at this level have a learning and deductive component of user profiles to help a user who cannot formalize a query or specify a target for a search. DiffAgent (Carnegie Mellon University) and Letizia (MIT) are examples. Agents at this level are referred to as learning agents or truly intelligent agents.

COMPONENTS OF AN AGENT

Intelligent agents are computer programs that contain the following components:

• Owner. User name, parent process name, or master agent name. Intelligent agents can have several owners. Humans can spawn agents, processes can spawn agents (e.g., stock brokerage processes using agents to monitor prices), or other intelligent agents can spawn their own supporting agents. .
• Author. Development owner, service, or master agent name. Intelligent agents can be created by people or processes and then supplied as templates for users to personalize.
• Account. Intelligent agents must have an anchor to an owner's account and an electronic address for billing purposes or as a pointer to their origin.
• Goal. Clear statements of successful agent task completion are necessary, as well as metrics for determining the task's point of completion and the value of the results. Measures of success can include simple completion of a transaction within the boundaries of the stated goal or a more complex measure.
• Subject description. The subject description details the goal's attributes. These attributes provide the boundaries of the agent, task, possible resources to call on, and class of need (e.g., stock purchase, airline ticket price).
• Creation and duration. The request and response dates requested.
• Background. Supporting information.
• Intelligent subsystem. An intelligent subsystem, such as a rule-based expert system or a neural computing system, provides several of the characteristics described above.