EXPERT SYSTEM 5)
← Back
A computer program able to resolve some problems in a way comparable to an expert or a group of experts.
An expert system includes a data bank and rules to use them. The rules are constructed by "knowlegde engineers" who extract the knowledge considered relevant from conversations with human experts and transform them into "production rules" usable in a computer to compare and interconnect data in significant ways, in order to produce useful conclusions. This is obtained through an "inference engine" which establishes coordinated general ways to apply the production rules to data corresponding to any specific case. This leads to the shaping of a decision tree.
The classical expert systems are thus typically algorithmic, but they ultimately depend from human expertise. Nevertheless, the more advanced connection machines permit the use of inductive programs, using "… a purely logical approach to find abstractly defined regularities in the data, irrespective of subject-matter" (M. BODEN, 1991, p.186).
Thus: "A learning program that uses this logical strategy can structure a classification-by-property conceptual space in the most economical way, and can find the shortest pathway for locating examples within it" (Ibid).
The rules collected in an expert system may also represent the consensus making process between different objectives (See ELF Supplement – New Scientist Vol 138 – n* 1870 – 04.24.93, p.7).
ELF Company applied the following features in its GRANDPUITS refinery in France, for refinery automation.
The proposal was to ensure the quality of production, as well as the long life of the unit, but also taking in account the action required to run the plant on a day-to -day schedule. This included:
"… on the spot analysis of the behavior of each part of the unit;
" decision-making, in order to create a control strategy taking into account the time factor of the actions;
" more detailed control procedure with the commands routed to the controllers through out the network.
"This required that the system be fully integrated in the real-time environment" (Ibid).
Expert systems are designed within the technical perspective and so, are weak or useless for the organizational or societal perspective, and still more so for the personal perspective (I.I. MITROFF and H.A. LlNSTONE, 1993, p.105).
Categories
- 1) General information
- 2) Methodology or model
- 3) Epistemology, ontology and semantics
- 4) Human sciences
- 5) Discipline oriented
Publisher
Bertalanffy Center for the Study of Systems Science(2020).
To cite this page, please use the following information:
Bertalanffy Center for the Study of Systems Science (2020). Title of the entry. In Charles François (Ed.), International Encyclopedia of Systems and Cybernetics (2). Retrieved from www.systemspedia.org/[full/url]
We thank the following partners for making the open access of this volume possible: