26 Expert System for Development

T. Radha

epgp books

 

 

 

 

Introduction

 

This module about the importance of various expert systems used in agricultural domain. Present expert systems play the role of agricultural engineer and provide users the different methods of diagnosis and treatments. In agriculture, expert systems unite the accumulated expertise of individual disciplines, e.g., plant pathology, entomology, horticulture and agricultural meteorology, into a framework that best addresses the specific, on-site needs of farmers.

 

GENESIS OF EXPERT SYSTEMS

 

ES technology has resulted as a spin-off from AI. AI is a specialized field of computer science which has been studied since the mid- 1950s. Its primary aim is to make machines emulate human intelligence. Mockler and Dologite (1992) define AI as “the capability of a device, such as a computer, to perform functions or tasks that would be regarded as intelligent if they were observed in humans”.

 

The initial research into AI seemed very promising but this was followed by many disappointrnents, as the research was followed by many unsuccessful attempts at applying AI to real world problems. Because AI is aimed at solving the more general and all-encompassing problems, difficulty with AI systems often arise because more than just specialized skills and knowledge is required to solve many general problems. In fact, the common knowledge that most of us have and use extensively on a day+-day basis for decision making tends to be the area most lacking in AI systems and therefore the dominant reason for their failure.

 

This situation has resulted in some strange reasoning by AI systems and is the main reason that AI as a source of general knowledge is still in the research labs and is still far in the future. As mentioned above, ES are a spin-off of AI and can be thought of as the more practical and successful implementation of AI technology and experience.

 

These “mini AI” applications are more successful because they operate in very narrowly defined subject areas or knowledge domains.

 

By keeping the subject area narrow, adequate levels of knowledge can be encoded into the knowledgebase [i.e. “collections of expertise or expert knowledge” (MocklerDologite 1992)l of an ES application and then used to solve specific problems. In other words, ES technology has been successful because, unlike its precursor (AI), it works within specific rather than general application areas.

 

An Expert system is software application that attempts to reproduce the performance of one or more human experts . Expert systems are mostly based on a specific problem domain, and are a traditional application of artificial intelligence . The expert system is used to behave like a human expert to solve the problem with the help of pre-set conditions in the software application.

 

A wide variety of methods can be used to stimulate the performance of the expert, which are 1. The creation of knowledge base which uses knowledge, and

2.   A process of gathering that knowledge from the SME and codifying it according to the formalism, which is called knowledge engineering.

 

Expert system may or may not have learning components but a third common element is that once the system is developed it is proven by being placed in the same real world problem solving situation as the human SME,typically as an aid to human workers or a supplement to some information system.

 

As a premiere application of computing and artificial intelligence , the topic of expert systems has many points of contact with general and systems theory, operations research, business process reengineering, applied mathematics and management science.

 

Expert system , particularly in agriculture sector can be used effectively to provide proper advise to the farmers in the area of nutrition management, pest control, selection of crop based on soil and water availability and many more.

 

Importance of expert system

 

The complexity of problems faced by farmers are yield loses, soil erosion, selection of crop, increasing chemical pesticides cost, pest resistance, diminishing market prices from international competition, and economic barriers hindering adoption of farming strategies.

 

The farmer may not become expert manager of all these aspects of farming operations. On the other hand , agriculture Expert system post gradute diploma in agricultural extension management (PGDAEM) researchers need to address problems of farm management and discover new management strategies to promote farm success.

 

Numerical methods have failed to provide better solution because understanding about crop systems are quantitative based on experience and cannot be mathematically represented. Expert system are computer programs that are different from conventional computer programs as they solve problems by minimizing human reasoning processes, replying on logic, beliefs, rules of thumb opinion and experience. The experience and knowledge of scientists will be used todevelop expert system on various issues of agriculture, which in turn will provide advisory support to the farmers .

 

Disciplines In agriculture, expert system are capable of integrating the perspective of individual disciplines such as plant pathology, entomology, horticulture and agricultural meteorology into framework that best address the type of ad hoc decision making required of modern farmers.

  • Expert systems can be one of the  most useful tools  for accomplishing the task of providing growers with the day-today integrated decision support needed to grow their crops.
  • An expert system contains knowledge about a particular field to assist human experts or provide information to people who do not have access to an expert in the particular field.
  • A database programme retrieves facts that are stored, while an expert system uses reasoning to draw conclusions from stored facts. Expert systems act as intelligent assistants to human experts, and assist people who otherwise might not have access to expertise .

 

Although both expert systems and database programmes feature the retrieval of stored information, the two types of programmes differ greatly. In medicines, for example, a database programme might be useful for enumerating the symptoms of various illnesses; while an expert system might help to diagnose illness, determine its causes and suggest programmes of treatment. Database programmes contain knowledge is only declarative knowledge.

 

Since a database programme cannot draw conclusions by reasoning about the facts in its domain, the users of a database programme are expected to draw their own conclusions .In contrast, expert systems contain expertise, consisting of both declarative and procedural knowledge , which allows them to emulate the reasoning processes of human experts .

 

Components of an expert system

 

A variety of techniques are used to create expert systems. They differ as widely as the programmers who develop them and the problems they are designed to solve. However, the principal components of most expert systems are

 

1.Knowledge base

2.Inference engine

3.User inference

 

1. Knowledge base

 

Knowledge base contains both declarative knowledge (facts about objects, events and situations) and procedural knowledge (information about causes of action ).

 

Depending on the form of knowledge representation chosen, the two types of knowledge may be separated or integrated . Although many knowledge representation techniques have been used in expert systems, the most prevalent form of knowledge representation currently used in expert systems is the rule –based production system approach . in a rule – based system, the procedural knowledge, in the form of heuristic if –then production rules, is completely integrated with the declarative.

 

2. Inference engine

 

Simply having access to great deal of knowledge does not make a person an expert: one must know how and when to apply the appropriate knowledge. Similarly , just having a knowledge base does not make an expert system intelligent . The system must have another component that directs the implementation of the knowledge. The element of the system is known variously as the control structure, the rule interpreter , or the inference engine.

 

The inference engine decides which heuristic search techniques are used to determine how the rules with knowledge base are to be applied to the problem . in effect , an inference engines runs an expert system, determining which rules are to be invoked, accessing the appropriate rules in the knowledge base , executing the rules and determining when an acceptable solution has been found.

 

Since the knowledge is not intertwined with the control structure, an inference engine that workswell in the expert system may work just as well with a different knowledge base, thus reducing expert system development time.

 

User interface

 

Interface is a collection of operations grouped under a single name. Even the most sophisticated expertsystem is worthless if the intended user cannot communicate with it. The component of an expert system that communicates with the user is known as the user interface.

 

The communication performed by the user interface is bidirectional. At the simplest level, one must be able to describe the problem to the expert system, and the system must be able to respond with its recommendations . In practice, a user interface generally is expected to perform additional functions. For example, the user may like toask the system to explain its reasoning or the system may request additional information about the problem from the user.

 

Although the designers of expert system generally have a great deal of experience with computers, the intended users of expert systems are frequently computer novices.

 

Functions of expert systems

 

The main function of an expert system is to mimic expertise and distribute expert knowledge to non-experts. The rapid development of Internet technology has changed the way that an expert system can be developed and distributed, but the distribution of expert systems to a large scale of end users can be challenging in terms of both effectiveness and efficiency. From a knowledge transfer perspective, it is argued in this paper that an expert system application is a knowledge transfer process in which expert knowledge is captured by a computer system (ICT) and delivered to non-expert recipient. It is believed that looking at expert system applications from a knowledge management point of view could benefit researchers and practitioners working in either the expert system or knowledge management domains.

 

The traditional knowledge transfer approach is frequently criticized for its rigidity and bureaucracy because all knowledge transfer activities solely rely on the face-to-face communications from a centralized extension agency down to local recipients.

 

After a few layers of people-to-people contact, knowledge can be easily lost or distorted. At the same time, knowledge transfer cannot achieve a great deal of efficiency by limited source when dealing with many users.

 

Sharing knowledge is power” (Liebowitz, 2001). This is especially true when agricultural expert systems are introduced to farmers in the developing world. In many developing countries the agriculture sector remains the largest employer in the country and agricultural productivity is one of the major concerns for the country‟s economy.

 

Agricultural knowledge extension is seen as an effective solution for improving agricultural productivity. The word “extension,” in this context, derives from an education development in England during the 19th century when Oxford University and Cambridge University attempted to serve the rapid expansion of educational needs from society. It was called “university extension”.

 

In the early 20th century, the word extension was applied to describe the transfer of knowledge and technology to serve the needs of rural development by American land-grant universities (Jones &Garforth, 1997).

 

The rapid development of ICT, such as agricultural expert systems, brings new opportunities to the agricultural extension methodology (Rees, Momanyi&Wekundah, 2000). From a knowledge transfer perspective an agricultural expert system is a knowledge transfer medium, through which advanced knowledge is encoded and transferred to a recipient, who can learn and benefit from the knowledge transferred.

 

Expert system applications may involve a large number of users with diversified application scenarios. For example, a web-based expert system project in China can have potential impact on millions of farmers. Because of the large scale of the application, the application scenarios may vary in accordance with specific local farming conditions.

The success of large scale expert system projects in the agriculture sector faces a number of challenges:

 

1.the large knowledge gap between the knowledge provider and the farmers, as sometimes, poorly educated farmers are not able to absorb the knowledge delivered to them nor follow the advice provided by the system;

2.the sheer number of users involved in using the expert systems;

3.physical distances between the knowledge provider/knowledge engineer and end users;

4.complex and diversified application contexts.

 

EXPERT SYSTEMS IN AGRICULTURE

 

We have all read about “management oriented farming” and “the systems approach” as well as a variety of research thatrecommends “optimal” management under a variety of circumstances. The practical problem with this philosophy andwith this available research knowledge is that there is often too much information (i.e. “information overload”). Thisproblem often leads to decisions which produce less than optimal results because of data complexity and time constraints.

 

As Comeau and Goit (1988) point out “agricultural problems are often multi disciplinary and very complex in nature”. ES in agriculture, therefore, offer great promise because both currently available and future knowledge can be organized within a knowledgebase, as described above, and used within the problem solving process.

 

This encoding of knowledge can then allow for implementation of the “management oriented farming” philosophy by producing decision support systems which leverage management’s input into agricultural production systems.

 

ES in agriculture help people to make complex decisions about agricultural resource systems more effectively and more timely. Without this technology, many people, often without the desired level of experience or expertise, are forced to make decisions using incomplete information. By helping people to consider all of the relevant information and by assimilating this information into an understandable format, ES assist people in the making of environmentally sound and economically viable farm management decisions. ES technology, therefore, offer the potential to provide a necessary link between both research information and human expertise, and the practical implementation of this knowledge 

 

Agricultural production has evolved into a complex business requiring the accumulation and integration of knowledgeand information from many diverse sources. In order to remain competitive, the modern farmer often relies onagricultural specialists and advisors to provide information for decision making. Unfortunately, agricultural specialistassistance is not always available when the farmer needs it. In order to alleviate this problem, expert systems wereidentified as a powerful tool with extensive potential in agriculture

 

One of the advantages of employing expert system is its ability to reduce the information that human users need toprocess, reduce personnel costs and increase output. Another advantage of expert system is it performs tasks moreconsistently than human experts. Some diagnosing expert systems depend on the ability of an end user to understandabnormal symptoms of the plant and to convey these symptoms through a textual dialogue. Depending on the user‟slevel of understanding of the abnormal observations, the expert system can reach the correct diagnosis.

 

If, however, theend user interprets the abnormal observations in a wrong way and chooses a wrong textual answer to a presentedquestion, then the expert system will reach a wrong conclusion

 

Characteristics of Agricultural Expert System:

  •  It simulates human reasoning about at problem domain, rather than simulating the domain itself.
  • It performs reasoning over representations of human knowledge
  • It solves problems by heuristic or approximate methods

APPLICATION AREAS OF AGRICULTURAL EXPERT SYSTEMS

 

l)   Crop Management Advisors

2)  Livestock Management Advisors

3)  Planning Systems

4)  Pest Management Systems

5)  Diagnostic Systems

6)  Conservation/Engineering System

7)  Marketing Advisory Systems etc.

 

Advantages and Disadvantages of Expert system Advantages

  • Expert system are useful in many aspects and ready to use by end user as advisory system
  • Provides consistent answers for repetitive decisions, processes and tasks
  • Holds and maintains significant levels of information
  • Encourages human expert to clarify and finalise the logic of their decision making

Disadvantages

  • Lacks common sense needed in some decision making
  • Cannot make creative responses as human expert  would in unusual circumstances
  • Domain experts not always  able to explain their logic and reasoning
  • Cannot adopt to changing environments, unless knowledge base is changed.

 

The discipline of agricultural science is decision oriented but the understanding of resource systems such as agricultural production is poor compared to, for example, mechanical or electrical systems. This uncertainty can be captured and dealt with as effectively as possible using expert experience and knowledge within an ES development environment. ES technology is becoming more widespread as the number of constraints imposed on agricultural production systems continue to increase. For this reason, agricultural research is an important part of developing and improving ES over time because the knowledge generated is used to improve our understanding of agricultural processes.

 

CONCLUSION

 

Expert Systems can be of great help to the farmers as well as the researchers. Their efficiency of diagnosing the rightdisease and treatment can enhance the productivity and reduce the losses. Expert systems and decision support systemsare widely used in developed countries. the need of expert systems in agriculture andavailability of various expert systems in various countries. The need of expert systems for technical informationtransfer in agriculture can be identified by recognizing the problems. But most of the expert systems are in Englishlanguage. By developing an expert system in agriculture in a mother tongue of a farmer, helps him/her to know thefacts and truths in increasing the production

you can view video on Expert System for Development

 

References and Web links