![]() |
|||
![]()
|
![]() |
![]() |
![]() |
2.3. FACTORS INHIBITING EXPERT SYSTEM DEVELOPMENTWhile some of the factors that inhibit expert systems development in government are similar to those in business, others are unique to the circumstances surrounding government activities. The major impediments to governmental development of expert systems include cost, knowledge engineering difficulties, technology orientation, time frame of government officials, voter attitude, and competing activities for limited funds. Cost. The costs of developing expert systems can be high, depending on the complexity of the domain, availability of experts, etc. Unfortunately, many, if not most, governmental agencies do not have access to the large research and development budgets of some business organizations. The Internal Revenue Service (IRS), for example, began an artificial intelligence development program in 1983. Because the IRS had no previous experience, a considerable investment in training and technology was required, and the annual budget now exceeds $5 million (Rogers and Beckman, 1990). Fortunately, the cost of the necessary computer hardware and software has been falling and expert system shells and tools are providing quicker development. Furthermore, given the size of government, significant economies of scale can often be achieved by government organizations, i.e., more people may use an expert system than in the private sector where several organizations perform the same function but do not share an expert system for competitive reasons. Knowledge engineering difficulties. In the public sector, the scarcity of knowledge engineers is a major factor inhibiting expert systems development. In addition, in some areas of government where an expert system could be most useful, communication between knowledge engineers and domain experts can be quite difficult, e.g. diplomats in far-flung regions. Also, the need to update systems for changing environments in such domains as diplomacy could be onerous. Technology orientation (or lack of it). For an organization to effectively develop and use expert systems, its people must be willing to adapt to new ideas and new technologies. Organizations without a technology orientation are more likely to experience expert systems development failure. Predictably, many of the governmental agencies that have already successfully implemented expert systems are those that had a forward-looking technology attitude even before the expert systems development was begun, such as military and defense departments and space exploration agencies. Time frame of government officials. One of the characteristics that defines public institutions is the short time perspective of its top managers, sometimes averaging as little as 18 months. This relatively short time horizon that shapes governmental unit culture is a result of political constraints. One of the casualties of this short time frame is technological innovation and investment, especially those like expert systems that take a considerable amount to time to develop. Because of the pressure to see results immediately, more short- term, highly visible projects have a higher priority. Short-term voter view. Expert systems development is not accomplished overnight. Because governmental bodies are altered as voter' preferences change, voters' evaluation of an expert system project may be crucial to its long-term success. Unfortunately, voters tend to have a short-term view. As today's society becomes accustomed to almost instant information and feedback, voters are likely to adopt an even shorter term focus. Therefore, expert systems are most appropriate for governmental tasks that are not radically transformed by a change in elected officials. Competing activities for limited funds. Taxation and budgetary allocatoins provide the monetary resources for public institutions to perform their activities and accomplish their mission and goals. Most information resource management (IRM) expenditures in the public sector are lower than in the private sector, as low as 1 or 2% of budget. Because of limited resources in the public sector, many projects are competing for them. In fact, most dollars spent on information technology go into maintaining existing applications rather than developing new systems, including expert systems. Hence, the non-committed resource available to develop expert systems can be severely limited. In analyzing where states are choosing to expend resources, the major portion of them are building their infrastructure to improve transaction processing and communication, especially with the states' citizens. If the public sector follows the pattern of information technology implementation that the private sector has, only after their infrastructures for transaction processing and management information systems have been well developed will they recognize the usefulness of expert systems. 3. APPLICATIONSWhile expert systems development has occurred in a wide variety of government agencies, only five principle domains are addressed in this section. Some areas involve security and defense and therefore have limited publicly available information. Other areas are so specialized that they do not provide useful examples to promote expert systems development in other government departments or countries. The five areas that are discussed below are environment and pollution, construction and planning, law and law enforcement, social services, and taxation. These areas are particularly suited to expert systems development because they are highly regulated tasks, are important to voters, and are generally applicable to most nations, states, and localities.
|
![]() |
|
Use of this site is subject certain Terms & Conditions. Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details. |