EXPERT SYSTEM – NATIONAL INFORMATICS EXPERT SYSTEM TECHNOLOGY By Javed Ashraf
Section – A: Present & Future Perspectives
EXPERT SYSTEM – NATIONAL INFORMATICS
EXPERT SYSTEM TECHNOLOGY
Prof. S.J. Ashraf*
1. INTRODUCTION
It has been said that non experts outnumber experts in any particular field by a ratio of 400 : 1 ; perhaps, this is truer of the developing countries. One of the existing reasons for this disparity is the lack of motivation of learning systems which are not significantly oriented to expertism in a given field. The momentum of the progress in the technologically advanced countries is such that the technological gap between the developed and developing countries is widening overall. Developing societies are being left behind because of their slow rate of absorption of technology. Before they can properly digest the inventions of the day, many more new inventions have been added. Consequently this gap penetrates right to the level of expert-ism of the developing countries. This expert ism is the main vehicle to sustain momentum in the transfer of technology to the developing/threshold countries.
One of the information technology legend means 6 is known as Expert System. It has been characteristically designed to enhance the communication of knowledge right up to the level of the end-user in the form of required expertise. This is a step forward towards disintegrating the existing expertise and experts. Experts and expertise are identified differently. However the researchers are devoted at full swing to recognise the new role of human experts. Presently the technology gives users the choice of decision according to the requirement as to whether he wishes to hire expertise or experts.
The implications of this technology will influence the society as a whole, its norms, its values, and the functional needs of productivity; hence the organizational set-up will be able to channelize expert ism in all vital fields of human endeavour. Nations competing in technology may look forward to the expertise in a suitable and acceptable form. Nevertheless, the acquisition of this technology could bridge the widening gaps and can move a step ahead towards the notion of “expertise self-sufficiency” in fields of national interests.
It is proposed to review how this higher technology area can help to overcome the dearth of experts by acquiring the relevant expertise for complicated problem handling in the cross section of our society.
Through experimental successes attempting efforts geared to the development of Artificial Intelligence (AI) has successfully emerged2 the emulated production of human-like expertise which may be called pseudo-expertise. More or less in the last three decades, laboratory curiosities of applied artificial intelligence have evolved into technical and commercial efforts. Marvin Minsky, Professor of Artificial Intelligence of Technology at Massachusetts and one of the founders of AI has stated that “artificial intelligence is the science of making machines do things that would require intelligence if done by men”,
One of the most economically rewarding areas of AI implementation today is that of the “Expert System”, also called the knowledge-based system. This system employs computers in ways that differ markedly from the conventional data-processing applications, which are algorithm-oriented, and instead they can:
– solve very difficult problems as well as or better than human experts;
– reason heuristically, using what experts consider to be the effective rule of thumb, and then interact with humans in appropriate ways, i.e. in the natural language;
– manipulate and reason about symbolic descriptions;
– can function with data which may contain errors, using uncertain judgement rules;
– can contemplate multiple competing hypotheses simultaneously;
– can explain why they are asking a question; –
– can justify the conclusion.
These ”expert systems” are to achieve the knowledge-engineering objectives which embody, inter-alia, the following capabilities:
– Efficiency; in order to minimize cost and response time;
– Modifiability; easy to add to or delete from the knowledge embodied;
– Simplicity; keeping the technology as ample as possible;
– Understandability; to help promote user acceptance;
– Correctness; provably correct conclusions, consistency and completeness of the knowledge base.
In a short and simple way, “Expert Systems” are computer programmes which parallel human reasoning. Unlike the conventional systems, which deal with rigidly structured procedures, expert systems can weigh evidence, cope with judgement and uncertainty and can explain their recommendations. They can also be improved, as knowledge is refined. Expert systems contain a knowledge base, which embodies human experts’ knowledge by means of a question-and-answer session. All this is because the computer can embody data, knowledge and inference mechanism which a single person, howsoever capable, just cannot possess.
Human expertise is essential to any organizational functioning. “Expert Systems” provide the means of dissemination of knowledge in a comprehensible and cost-effective manner, thus leaving the specialists free to further develop their own expertise.
2. IMPORTANT ADVANTAGES
Expert systems are reliable. Humans (even experts) can, on occasion, forget details which may be vital. Expert systems do not fail to ask relevant questions or absent-mindedly ignore important factors which may be up for discussion.
Expert systems, moreover are consistent. Given the same input, the Expert system consistently reaches the intended conclusions.
Expert systems are always available and can be interrogated by many users at once. Human experts often have to travel to deploy their knowledge, and are restricted by time and other business pressures. Expert systems can be copied and distributed whenever there is a suitable computer. Unlike human experts they can never take holidays and can be made available for 24 hours a day without intermission.
3. TECHNOLOGY POTENTIAL CHARACTERISTICS
Developing economies and their societies are mainly Lobar intensive (shop floor), and the organizational set-up is office-worker (white collar) oriented. The earlier phase of information technology faced considerable hesitation due to the fear of being job killers. The add-on features of this technology, instead of acquiring the automation of clerical activity, is now targeting at the automation of the intellectual components of task performance. By successively acquiring this technology, the developing countries should be ready to make available the expert consultancies services at large.
In order to avoid waste in money and efforts, the new technology, should be acquired along well-defined and prescribed lines, and should be dedicated to the furtherance of national interests, As the effects of the new technology acquiring process would be felt along a wide spectrum of the society as a whole, the objectives must be set and given sets of priorities, A mechanism for monitoring the tracks of acquisition, prescribing technology standards, the suitability of products developed for marketing and sales, and checking the flooding of the market with sub-standard or unmarketable, goods, has to be evolved. This approach embodies a check similar to the one, which is applied in the printing industry. As the impact of the new technology is bound to be tremendous, it needs checks and balances in a technology based society.
4. SOCIAL IMPACT
The debate on the social implications of computing has been going on for a long time. With developments in information technology, new possibilities are emerging and new threats are being perceived. The newly emerging ‘expert systems’ make a claim upon their capability of human-like style of computing. In a recent report of a specialists group on Expert Systems of the British Computer Society, it emerged that in abstraction the expert systems are revolutionary and evolutionary – revolutionary in handling empirical knowledge that might be incorrect and incomplete, and evolutionary in that the knowledge is explicit and separate from the programmes. The Expert System’s incompleteness is offset by its intelligibility.
Some of the other implicit attempts of the Expert Systems are:
– (their attempt) to mechanize the un-mechanize-able. Is an attempt undesirable? Is it inevitable?
– when the bulk of Expert System will flood the market, quality standards are bound to be question-marked. This would go to affect the services and the type of services the society would be getting from the human experts. Some of the vital service areas would be in the field of medical diagnoses, and follow-up treatments. The acceptance limits of Expert systems for complex decisions and delicate judgements are to be fixed.
In such situations there could be manifold impacts. Medical dilemmas could probably be solved by automatically applying a set of rules, not necessarily of a medical expert, on the specialists’ medical knowledge available with the machine base. People will tend to concentrate their attention elsewhere assigning such responsibilities to a machine which would be unable to take account of changing social conditions, moral values, and advances in medical technology affecting the economics of the health care.
One of the potential, dangers of the Expert systems resides in abdication of responsibility by human experts to the computer-based Expert systems usage and in the tacit and uncritical acceptance of the conclusions offered by the autonomous Expert systems.
Sociologists are rather wary about the way in which the new technology will make its impact felt upon society. One of the chief impacts will be that, consequent upon substantial reduction in manual work and reduced manpower the number of leisure-hours will increase. This, in itself, might lead to increased or decreased creativity, depending upon the society as a whole or the individual concerned. This might, according to some sociologists, evolve bad mankind towards unknown “dehumanizing” tendencies. Another harmful impact might be upon the higher creative skills, upon diversity which is the spice of human life, and the birth of a uniformly-oriented society in which human beings might act like the robots they themselves have designed.
5: EXPERT SYSTEM APPLICATIONS
The applications of the Expert Systems commenced as early as 1965. Some of these are:
1. DENDRAL: the first commercial expert system, for organic compounds analysis, using spectroscopy.
2. META-DENDRAL: the first successful learning system.
3. MYCIN: Medical information logic, to diagnose blood and meningitis infection and advise physicians about antibiotic therapy.
4. EMYCIN: Used successfully by structural engineers, who have specialized in structural mechanics, to identify the best strategy for using complex computer simulation?
5. TEIRESIAS: the first programme to provide explanation for putting queries of a particular nature.
6. INTERNIST: Capable of entertaining an adequate number of simultaneous hypotheses, not possible to the human mind.
7. CADUCEUS: was equipped with a model of the human body and is capable of making inferences based on the interaction of different organs, and can follow the progress of disease symptoms over time.
8. PROSPECTOR: Consultant system for mineral exploration.
9. INTELLECT: is intended to be a front-end interface to information-retrieval applications in areas such as finance, marketing, manufacturing, and personnel. It can resolve ambiguity by assigning preference rating to different paraphrases.
10. HAM-ANS: A German language interface, specially tailored to database access in a number of areas.
11. LOGOS: A machine translation system, working in partnership with human translators.
12. ALPS: Also an interactive translation system, finding unknown word or words with ambiguous meaning. It utilizes several reference dictionaries and builds up a separate dictionary for each document.
13. EPISTLE: provides business letter-writing support (spelling, grammar, and style checking) for office workers.
14. ARPA: Designed to support HEARSAY-I, HEARSAY-II speech-understanding systems.
15. DECTALK: It converts, alphanumeric text into human quality speech.
16. FEASA: “Finite Element Analysis Specification Aid”, to help engineers convert a real design problem, efficiently and accurately, into representation needed to use the NASTRN, a complex suit of programmes in use by BAe design engineers.
17. SACON: To advice engineers how to use a complex structural analysis programme.
18. XCON: generates a configuration and detailed plan upon how a selection of computer components should fit together.
19. KBTA: designed to find a hardware list for engineers in the process of test programmes and developments.
6. FEASIBILITY OF APPLICATIONS FOR EXPERT SYSTEMS EXPOSURE
The present-day industrialization, which is price and quality oriented, high-lights the need to integrate and coordinate knowledge about all phases of product systems. A variety of applications falls under the scope of applied AI and expert systems. The guidelines which can help in selecting a problem for knowledge-engineering exposure and success of application are :
— seeking a problem that expert can solve in a short span of time.
— choosing a problem whose solution requires symbolic reasoning.
— preferring a high-value problem.
— ruling out problems where different experts disagree about solution correctness.
— ruling out where initial problem solution does not lie within the scope of the expert’s total knowledge.
— selecting an initial class of problems to solve which requires only a subset of knowledge.
— identifying some training problems and collecting the experts’ problem-solving protocol,
— building a knowledge-base that contains an explicit and a declarative representation of the experts concepts and heuristic reasoning rules.
— developing an initial expert system to solve the training problems as the expert does.
— asking the expert to review the system’s solution and its lines of reasoning.
— augmenting the system to accommodate the expert’s critique.
— applying the system to more training cases and to augmenting its knowledge-base incrementally.
— evaluating the system’s performance on novel test cases.
7. POTENTIAL APPLICATIONS OF EXPERT SYSTEM
Use of Urdu in Computers
Dr Michael Agi (Gassellschaft fur Mathematik and Datenveiorbeiting m.b.h, Bonn), has given an account of aims of living human languages. Two of the basic aims are :
1. They must support the communication requirements and communication behaviour of a society.
2. As the main carrier of human history, culture, and civilization, the language must be able to cover all intellectual and social spheres of the
society.
Languages which cannot fulfill these two requirements will sooner pr later fade away or will be partly replaced by other living languages.
Japanese argue for a means to allow Japanese programmers to interact with computers, using their everyday language. The written form of the Japanese language is not amenable to programming. Programmers would be able to work much more efficiently, if they could, in their own language.
This argument also applies to Urdu. Putting it another way, one might assert that Urdu has to accept the modern challenges faced by a language which can fulfil the man vs man, and man vs machine communication-requirements, For the use of Urdu in computers, there are mainly two areas: an Urdu language processor and Urdu text-processing for linguistic and lexical analysis applications.
Urdu script has its own characteristics. Some are quite different from English (sinistral type of language) writing. Standardization of character sets and its associated codes (digitization; perhaps different from that existing in ILM-E JAFAR).
Some of the bottle-necks which need re-solution are :-
— character display and printing techniques
— writing orientation
— position sensitivity
— complexity of calligraphic shapes
— use of short vowels .
— bilingual! test writings (including dextral and sinisterly ambidextrous texts).
These problems need, at large, the quest of linguistics for domain expertise, and expert systems knowledge engineers for an intelligibility of problems and problem-solving for general acceptance.
Agriculture
Expert systems can evaluate the risk of crop diseases and give advice to farmers together with expertise of agronomy. Data could be recorded for each field to enable the farmer to receive advice tailored to his local conditions.
Medicine
Allopathic treatment and diagnoses attracted the immediate attention of the Expert System pioneers, and since then it is on the course of continuous improvements. The methods of treatment of human diseases are different in local environments. They include Homeopathy, and ‘Unani’ medicine, and other folklore and empirical systems like Mexican, and local African
tribal systems. The Expert system application could be of interest in these fields for the diagnosis of a disease and follow-up treatment. Besides, a super master database of experience could be created for future improvement and research in the fields.
Waterlogging, salinity, soil erosion, and desertification
Several techniques are available for the amelioration of waterlogging and salinity–the products of a bid irrigation system. Each measure has its limitations. The Expert system can advise on the use of a technique in a particular region, giving a mosaic of its characteristics. It could also analyze the causative factors leading to waterlogging and salinity and help to programme the future line of action.
Soil erosion is the result of the depletion of the top soil. The Expert system can help to monitor the loss of arable soil at certain fixed intervals and provide data upon the use of techniques to be followed in a particular region.
Other applications
— heuristics control-system for military functions and strategic planning.
— fault diagnosis in control-processing (heavy industry), and communication networks and maintenance.
— resource allocation and logistics.
— legal expert system to provide legal expertise to a much broader spectrum of the community, concerning the following:
(a) Definitional law
(b) Normative law
(c) Case law
— expert systems applications in the Inheritance System in Islam.
8. FUTURE HOPES FOR TECHNOLOGY ADVANCEMENT
To accept the challenges of the coming decades, the lion’s share of economic power will lie in an effective management of information. With potentials of dominance over the world-scene, with economics and military pre-eminence at stake, research in advanced computing is being accorded priority throughout the world and this has developed into a race. A variety of approaches and efforts in this regard are converging in inventing and producing advanced hardware to overcome the present-day constraints and limitations of the existing computer systems. This newly developed computer hardware is capable of carrying out multiple processing at the same time and is, for this reason, comprehensive. The newly developed computer is a fifth-generation machine, the term being applied in an evolutionary sense. However, the name semantic of the computer varies from nation to nation.
Japan possesses limited natural resources. She depends upon the conversion of raw material into finished goods, and any slack-hand in her technological advancement will result in disastrous economic consequences for her. She has allocated enormous budgets in her efforts to develop information-handling through R&D work.
Major events of development with regard to computers in the US include the programmes planned by giant organizations such as DOD, IBM, General Motors and the cooperative efforts of other major companies through MCC in the areas of :
— architecture of advanced computers
— micro-electronic packaging and interconnection
— advanced software technology
— CAD for VLSI.
The Advanced Research Project Agency (ARPA) for defence is heavily supporting research in these areas.
The EEC countries have joined hands in this field and have formed bodies for cooperative efforts. ICL (UK), BULL (France), and SIEMAN (W. Germany) have pooled their resources and R&D efforts in implementing advanced knowledge-processing systems. ESPRIT, a project of the EEC countries along with some US organizations, has been set up to promote research and development in Information Technology.
A major UK programme for advanced information technology is the ALVEY Programme, which is aiming at developing necessary technologies for knowledge-base system in order to improve the country’s competitive position in this technology area. The Turing Institute is working to promote the education and knowledge of the public in AI related fields.
West Germany has launched a programme to bring industry and academia together in research projects in this field. In Hungary, the Institute of Computer Coordination has developed experience with Expert Systems and is sharing efforts with EEC and USSR. The USSR is organizing research through a plan with the cooperation of Bulgaria, GDR, Hungary, Czechoslovakia, Cuba, and Poland.
India recently has announced an ALVEY-type plan and has expressed her readiness to join the western nations in the competition to build a new generation of thinking computers.
9. CONCLUDING REMARKS
Some countries are engaged in the core research and inventions, while others are getting ready to absorb the new technology.
It is high time, considering the wide range of the applications of the Expert Systems, to make knowledge-engineering a part of the Country’s national policy. Only its incorporation within the framework of the national policy will endow upon this new technological field a positive direction. Fragmented efforts by individuals or private concerns will not be able to pave the way for a major break-through. Moreover, certain fields like agriculture, development of computational systems in Urdu, and so on, would best be served through sponsorship by the Government. The time-span from one generation of computers to another has now been squeezed to three years, and this fact alone attests to our absolute backwardness in this major field.
ACKNOWLEDGEMENTS
Thanks go to my organization BCCI-FAST, ICS, Karachi for their continuous encouragement and support. My colleagues of applied artificial-intelligence
********
MESSAGE: DEDICATED TO ANONYMOUS COMBATANTS OF KNOWLEDGE
مکتب علم الل مہد منل لحد Learning continues from birth to death
FOR PROMOTION OF LEARNEDNESS SHARE WITH FRIENDS ABOUT;
- WHETHER ‘BRAIN-POWER’ OR ‘HEART-IMPULSE’ PLINTHS ‘HUMAN GEN’? http://sunedu.wordpress.com/2012/09/17/whether-brain-power-or-heart-impulse-plinths-human-gen/
************
Related
No comments:
Post a Comment