Previous | Table of Contents | Next |
Barbara J. Haley and Kelly Hilmer
The World Wide Webs potential as a strategic information tool is limited by the time, costs, and frustration associated with Web use. Intelligent agents that facilitate searching, monitor information changes, and customize and order information according to individual needs help managers reduce information overload and more effectively use the wealth of information available on the Web.
Organizations that invest much time and effort in information systems that support unique management needs are increasingly integrating the World Wide Web (Web) into their IS portfolio. Although countless articles have espoused the Webs promise to provide information and commercial opportunities to the business community, several current limitations make the Web fall short of expectations.
Recently, intelligent agents have become practical solutions for addressing these limitations. The use of intelligent agents is important for Web developers, IS and business managers, and organizations interested in maximizing the benefits of the Web as an information tool while minimizing the time, costs, and frustration associated with Web use. After describing how the Web nominally supports management information needs, this chapter illustrates how five types of intelligent agents address the current Web weaknesses managers may encounter.
Managers are unique systems users because of the wide variety of tasks they perform, the diversity of information they require, and the dynamic nature of their decision-making environment. The literature on executive information systems, for example, illustrates that executives require special systems that:
Over the years, systems have evolved to address these requirements and provide executives with effective tools for accomplishing management tasks. The Web appears to be yet another innovative way to address such needs. A broad range of internal and external data is accessible from Web sites within an organization and from external sites located around the world. Managers can use the available graphical packages to navigate the Web without assistance. Web browsers also present information in creative ways, taking advantage of graphics, sound, and video.
Information overload, constant updating and reshuffling of information, and minimal support for novice users are some of the current problems that reduce the Webs value to managers. The lack of Web regulation and control combined with rapid growth have contributed to an abundance of information. Because this information changes constantly with little order or customization, it is difficult for managers to locate information efficiently, especially because they typically navigate this chaos with minimal or basic computer skills. As managers increase their Web use, organizations will become more concerned about these problems.
Intelligent agents (agents), also called software agents, are software programs that help find, organize, and present information that is custom tailored to a meet a managers needs. The programs are relatively autonomous; they are not attached to a particular software application and do not need a user for activation. Agents always are ready to perform a specified action according to preset user parameters. In addition, advanced agents learn trends and user preferences, and they can analyze information to support decision making.
Agents address a variety of problems. They are designed to save time, perform mundane tasks, fulfill requests tailored to specific needs, and manage data. To some people, these agents serve as personal data assistants, addressing diverse information requirements. But, unlike a human assistant who needs sleep and occasional days off, an agent addresses customized needs around the clock.
There are many types of agents, and a multitude of classification schemes have been suggested to help understand what intelligent agents can do. In general, agents vary in terms of three dimensions: agency, mobility, and intelligence. These dimensions respectively refer to the autonomy of the agents performance, the amount an agent traverses through networks, and how much an agent can learn or adapt to user requests. Agents can perform several functions, including:
Previous | Table of Contents | Next |