Academic journal article Competition Forum

Intelligent Software Agents and the Creation of Competitive Advantage

Academic journal article Competition Forum

Intelligent Software Agents and the Creation of Competitive Advantage

Article excerpt

EXECUTIVE SUMMARY

An agent is developed with empowerment to act on its own by making its own decisions and reacting independently to its problem domain. An agent will determine how capable it is to change its own course and respond to obstacles it meets while achieving its goal. An example of autonomy would be an agent's ability to act upon of varied level of requests and prioritize its response at each level. An intelligent agent may be able to ask questions and seek other intelligent response in order to modify its own course rather than blindly following a less effective path. The level of autonomy is dictated by its goal orientation where the agent determines how and where to satisfy its owner's request. Its ability to modify requests and respond to its environment is characteristic of its collaborative nature. Agents are flexible because their actions and reactions are not scripted to the degree that limits their ability to dynamically choose and sequence appropriate actions. Agents are self-starting when they can sense changes within their problem domain that require their action or reaction (Hess et al., 2000).

INTRODUCTION

Intelligence levels: the following table contains levels of agent' intelligence, a brief description of each level of intelligence, and an application example of each level of intelligence (Lee et al., 1997).

AGENT CHARACTERISTICS

Mobility is the degree to which agents travel through the network. Agents may reside on a client machine, the server, or travel among machines. Some agents present their credentials, like a password or user profile, to obtain access to services and data managed by agencies.

Capability refers to the goal for the intelligent agent and is defined by its intelligence level.

Task orientation ranges from accomplishing a single task to interaction with other agents, programs, or humans. The proactive nature of agents is to initiate goal-directed behavior resulting from their owner's direction or of their own selfdirected action. The reactive nature of an agent is its recognition and timely response to changes in within its problem domain (environment). An agent may contain a proactive nature, a reactive nature, or both proactive and reactive natures.

Agents, particularly those at level 2 or level 3 intelligence, may be expected to interact with human users. Agents may be designed specifically to characterize specific personality temperaments (extraversion and introversion, sensation and intuition, thinking and feeling, and perceiving and judging).

Temporal Continuity refers to a continuously running process that either "wakes up" to periodically complete its goals or is inactive and waits for a specific event.

Time driven and event driven characteristics may be viewed as a subset of temporal continuity because they relate to the continuous operation of an agent and determine the basis by which the agent becomes active. Event driven characteristics indicate a reactive nature of the agent. A time driven characteristic indicates a proactive nature of the agent.

The goal for many agents includes their ability to work in the background, meaning their task may provide information for other agents, programs, or users on a continuous basis, but with less priority than a foreground task. Other background agents are inactive until triggered by the specific event or elapsed time.

Only level 3 intelligence characterizes an agent as being able to learn, and most agents do not perform tasks that are level 3 tasks. Software wizards interpret users patterns of behavior and offer suggestions to the user on how they can perform their tasks more efficiently. Precede and succeed characteristics refers to the order in which one agent is spawned and executed within a larger set of tasks and activities. Order of tasks may change, due to the decision made by agents within their problem domain.

REASONS FOR USING AGENTS

Intelligence software agents are significant to business processes, goals, and strategies primarily due to information overload. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.