Agents for Handheld Devices

Agents are artificial intelligent programs that are capable of performing autonomous actions by controlling their actions. An agent may be able to discover and interact [Page 18]with other agents (including humans). Some agents may perform reactive actions through sensors. Such agents will react to changes sensed in the environment. Other agents do not simply react but are proactive, exhibiting goal-oriented behavior. Agents can be classified into four types: autonomous, communicative (socially able), reactive, and proactive. An agent may be a single agent system consisting of one of the four types of agents, or a multiagent system consisting of more than one of the four types. Agents can also be classified as reasoning agents, task-specific agents, communicative agents, information agents, collaborative agents, interface agents, and mobile agents.

Agents can be described in terms of the environment, abilities of sensors, built-in actions, built-in preferences, and decision-making mechanisms. Stan Franklin, the founder of the Institute for Intelligent Systems at the University of Memphis, defined this as a specification for describing an autonomous agent. The four types of agents can be considered as basic properties. Michael Wooldridge and Nicholas Jennings argued that other properties of agents include temporally continuous, learning, mobile, flexible, and character specific.

For educational purposes, agents have been used to assist instructors and learners in the teaching and learning process. This entry first describes the four main types of agents used in education and then discusses the application of mobile or handheld agents, a more recent advancement in agent technology, to teaching and learning.