Você está na página 1de 13

Knowledge based Systems

Unit 1

Unit 1
Structure Objectives 1.1 Artificial Intelligence (AI) 1.1.1 Components of AI 1.2 Expert Systems (ES) 1.2.1 Components of ES 1.2.2 Criteria for adopting ES

Knowledge based Systems

1.2.3 Advantages & Disadvantages of ES 1.2.4 Keys to successful ES development 1.2.5 Applications of ES 1.3 Neural Networks 1.3.1 Neural network process 1.3.2 Applications of neural networks 1.4 1.5 Robotics Exercise

Objectives This unit deals with knowledge based system like artificial intelligence, expert systems, neural network, robotics etc. At the end of this chapter, reader will be in a position for successful expert system development.

1.1 Artificial Intelligence (AI)


Artificial intelligence can be defined as a field of study that designs and develops machines capable of performing tasks that would require intelligence if performed by a human being. A more formal definition of artificial intelligence is that, it is a branch of computer science concerned with designing intelligent computer systems, ie; systems that exhibit the characteristics associated with intelligence in human behavior understanding, language, learning, reasoning etc.
Management Information System Block II Page No.: 1

Knowledge based Systems

Unit 1

1.1.1 Components of AI Broadly speaking, AI consists of (a) Expert systems (b) Robotics (c) Natural Language (d) Learning (e) Computer Vision (f) Perceptive Systems (g) Artificial Intelligence hardware (h) Neural Networks (a) Expert Systems: It is a computer program that attempts to represent the knowledge of human experts in the form of heuristics. (b) Robotics : They consist of computer controlled devices that help in AI applications. (c) Natural Language : This enables users to communicate with the computer in different languages and also enables the computer to check grammar and spellings. (d) Learning : This encompasses all the activity that enables the computer or other device to acquire knowledge in addition to what has been entered into the memory by its manufacturers and programmers. (e) Computer vision : This is to endow computers with the ability to recognize and identify objects and the context to which they belong. This also entails an ability to recognize shapes, features etc. automatically and in turn automate movement through robots. (f) Perceptive Systems : These use visual images and auditory signals to instruct computers or other devices such as robots. (g) Artificial intelligence hardware : This includes physical devices that help in artificial intelligence applications. Examples are hardware that is dedicated to knowledge based systems, neural computers used to speed up calculations etc.
Management Information System Block II Page No.: 2

Knowledge based Systems

Unit 1

(h) Neural networks : They are highly simplified models of the human nervous system that exhibit abilities such as learning, generalization and abstraction. These abilities enable the models to learn human like behavior.

1.2 Expert system (ES)


An expert system is a computer application that guides the performance of ill structured tasks, which usually require experience and expertise. Using an expert system, a non expert can achieve performance, which is comparable to an experts performance in that particular domain. An expert system is very similar to a decision support system, ie; both are intended to provide a high level of problem solving support to their users. But they differ in two major ways : First, a DSS consists of routines that reflect as to how the manager believes a problem should be solved, as well as the managers style and capabilities. An expert system on the other hand, offers the opportunity to make decisions that exceed the managers capabilities. The second but most important difference between DSS and ES is the ability of the expert system (ES) to explain its line of reasoning in reaching a particular solution. Very often, the explanation of how a solution was reached is more valuable than the solution itself. 1.2.1 Components of ES Expert system consists of 4 major parts (a) (b) (c) (d) user interface knowledge base inference engine development engine

Management Information System Block II

Page No.: 3

Knowledge based Systems

Unit 1

Instructions & Information

Solutions & Explanations

Knowledge

User

User Interface

User Interface

Knowledge Base

Problem Domain

User Interface

Expert and Knowledge Engineer

Management Information System Block II

Page No.: 4

Knowledge based Systems

Unit 1

(a) User Interface : The user interface enables the user to enter instructions & information into the expert system and also to receive information from it. The instructions specify the parameters that guide the expert system through its reasoning process. The information is in the form of values assigned to certain values. (i) Expert system inputs : The user can use 4 methods for input purposes : menus, commands, natural language and customised interfaces. (ii) Expert system outputs : Expert systems are designed to recommend solutions. These solutions are supplemented by explanations. There are two types of explanations : Explanation of questions and Explanation of problem solution Explanation of questions: The manager may desire explanations while the expert system performs its reasoning. Perhaps the expert system will prompt the manager to enter some information. The manager asks why the information is needed, and the expert system provides an explanation. Explanation of problem solution: After the expert system provides a problem solution, the manager can ask for an explanation of how it was reached. The expert will display each of the reasoning steps leading to the solution. (b) Knowledge base : The knowledge base contains both facts that describe the problem area and knowledge representation techniques that describe how the facts fits together in a logical manner. The term problem domain is used to describe the problem area. The popular knowledge representation techniques are rules and networks of rules. The rules in the knowledge base are usually coded in the form, IF X THEN Y, where X is a condition, Y is an action to be taken if the condition is true. All the rules contained in an expert system are called the rule set.

Management Information System Block II

Page No.: 5

Knowledge based Systems

Unit 1

The rules of a rule set are not physically linked, but their logical relationships can be established. (c) Inference engine : The inference engine is the portion of the expert system that performs reasoning by using the contents of the knowledge base in a particular sequence. During the consultation, the inference engine examines the rules of the knowledge base one at a time, and when a rules condition is true, specified action is taken. In expert system terminology, the rule is fired when the action is taken. Two main methods have been devised for the inference engine to use in examining the rules: forward reasoning and reverse reasoning. In forward reasoning, also known as forward chaining, the rules are examined one after another in a certain order. The order might be the sequence in which the rules are entered into the rule set, or it might be some other sequence specified by the user. As each rule is examined, the expert system attempts to evaluate whether the condition is true or false. In reverse reasoning, also known as backward chaining, the inference engine selects a rule and regards it a problem to be solved. Reverse reasoning is faster than forward reasoning because it does not have to consider all the rules and does not make multiple passes through the rule set. Reverse reasoning is appropriate when There are multiple goal variables There are many rules All or most of the rules do not have to be examined in the process of reaching a solution. (d) Development engine : The development engine is used to create the exert system and this process involves building the rule set. There are two basic approaches : programming languages and expert system shells.
Management Information System Block II Page No.: 6

Knowledge based Systems

Unit 1

(i) Programming languages : An expert system can be created using any programming language, however, two are especially well suited to the symbolic representation of the knowledge base, Prolog and Lisp. (ii) Expert system shells : An expert system shell is a ready made processor that can be tailored to a specific problem domain through the addition of the appropriate knowledge base. In most cases, the shell can produce an expert system quicker and easier than by programming. 1.2.2 Criteria for adopting ES The following factors should be taken into account if an expert system is required to be developed for a potential application area : (1) Availability : One or more experts are capable of communicating how they go about solving the problems to which the expert system will be applied. (2) Complexity : Solutions of the problems for which the expert system will be used is a complex task that requires logical inference processing, which would not be easily handled by conventional information processing. (3) Domain : The domain or subject area of the problem is relatively small and limited to a well defined area. (4) Expertise : Solutions to the problems require the efforts of experts. This means only a few people possess the knowledge, techniques and the intuition needed. (5) Structure : The solution process must be able to cope with ill structured, uncertain, missing, conflicting data and a dynamic problem solving situation. 1.2.3 Advantages / Disadvantages of ES As with all computer applications, expert systems offer some real advantages, but there are also disadvantages. The advantages can accrue to both managers and the organisation.

Management Information System Block II

Page No.: 7

Knowledge based Systems

Unit 1

Advantages of ES to managers Make better decisions by considering more alternatives Apply a higher level of logic in evaluating the alternatives More time to evaluate decision results Achieve a consistency in the decision taken

Advantages of ES to organisations An organisation that implements expert systems can expect Better performance for the organisation Maintain control over the organisations knowledge

Disadvantages of ES They cannot handle inconsistent knowledge They cannot apply judgement and intuition recognised as important ingredients of problem solving They cannot apply to unstructured / partially structured problems

1.2.4 Keys to successful ES development The following are the ingredients that should be incorporated into any system development : (1) Coordinate expert system development with the strategic business plan and the strategic plan for information resources (2) Clearly define the problem to be solved and thoroughly understand the problem domain (3) Pay particular attention to the legal and ethical feasibility of the proposed system (4) Fully understand both the users concerns about the development project and their expectations of the operational system (5) Employ management techniques designed to keep the attrition rate for developers within the acceptable limits.

Management Information System Block II

Page No.: 8

Knowledge based Systems

Unit 1

1.2.5 Applications of ES The business applications of expert systems are many. Some of them are as follows : Manufacturing Maintaining facilities Analysing quality and providing corrective measures Determining whether a process is running correctly or not Scheduling job shop tasks Selecting transportation routes Assisting with product design and facility layouts

Marketing Establishing sales quotas Responding to customer inquiries Assisting with marketing timing decisions Determining discount policies

Accounting and Finance Providing tax advice and assistance Helping with credit authorization decisions Selecting forecasting models Providing investment advice

Personnel Assessing applicant qualifications Giving employees assistance in filling out forms

General Business Assisting with project proposals Recommending acquisition strategies Educating trainees Evaluating performance
Page No.: 9

Management Information System Block II

Knowledge based Systems

Unit 1

Social Applications Counseling prospective students about study options Counseling expectant mothers about the various benefits available to them Advising buyers on the ideal home layouts

1.3 Neural Networks


A neural network is a program that models the pattern recognition capabilities of the human brain, thereby endowing machines with the ability to identify and classify faces, voices, pictures and written characters. Neural networks are programs that model the interconnections of human brain cells. Neural network can also be defined to be an array of interconnected processing elements, each of which can accept inputs, process them and produce a single output. The output may be passed on through a connection to another processing element or produced to the environment of the net. The processing elements may be actual hardware components or their operation may be simulated by software.

Management Information System Block II

Page No.: 10

Knowledge based Systems

Unit 1

1.3.1 Neural network process

O1

O2

PE 6 W46 W56 W47

PE 6 W57

Highest layers (output

PE 4 W14

PE 5

Intermediate layers (hidden layers) W35

W34 W15 PE 1 i1 W24 PE 2 i2 W25 PE 3 i3

Inputs

Lower layers (input layers)

A generic neural network is shown above. It is made up of interconnected processing elements (PE), which are self adjusting units that are firmly connected with other processing elements (PE) in the system. A processing element receives inputs from the user; each input has a certain weight assigned
Management Information System Block II Page No.: 11

Knowledge based Systems

Unit 1

to it. The weights influence the way the input is processed by the processing element, and the processing element is capable of automatically adjusting the weights of the inputs based on its past experiences. Each processing element processes the input and generates a single output signal to other processing elements in the network. The system observes the overall pattern of outputs generated by the processing elements and this pattern forms the basis for information, analysis and retrieval. Thus, unlike other computers, which are programmed to perform a task, neural computers are trained to perform a task. By controlling and fine tuning the type and flow of information among processing elements, the software gradually changes the connections between them, and in this way new information is learned. For example, one neural network learned to pronounce 20,000 words. 1.3.2 Applications of neural network The applications of neural networks include detection of credit card fraud, a diagnostic tool called Neuroscope, developed by researchers at IBM, runs on a PC and provides early warnings of failures in industrial machinery such as motors, cleaning tools and pneumatic tools, thus making predictive maintenance, a reality. Neural networks are extensively used in financial forecasting. By studying factors such as changing patterns in debt level, cash flow, earnings estimates and other financial variables that measure the health of an organisation. Neural networks also perform real estate appraisals by taking into account variables such as location, interior layout and building materials. The appraisals of the neural networks are more consistent and accurate than those of human appraisers, because the system also compares and studies data on similar houses in similar markets. While human beings are capable of taking into account only a few factors and analysing their correlations, neural networks take into account a large number of factors and perform a detailed analysis of the
Management Information System Block II Page No.: 12

Knowledge based Systems

Unit 1

correlation among hundreds of variables. This is the reason why they outperform human experts. Neural networks are expected to impact other business areas as well, such as looking for trend patterns in sales data, analysing strategy patterns within organisations, looking for duplicate names in mailing lists and making recommendations on credit or mortgage applications.

1.4 Robotics
Robotics is the integration of computers and industrial robots. A robot with artificial intelligence capability is an electromechanical manipulator, which is able to respond to change in its environment, based on its perception of that environment. This implies that such a robot has a sensory subsystem (usually machine vision) and is programmed to interpret what it sees, and to act upon the interpretation. For example, an industrial robot can manufacture one of many parts in its repertoire and manipulate it to inspect it for defects, recognizing very small departures from established standards. The most common industrial robot is a single mechanical arm that is controlled by a computer. The arm, called the manipulator, has a shoulder, forearm and wrist, and is capable of performing the motions of a human arm. The manipulator is fitted with a hand that is designed to accomplish a specific task, such as painting, welding, picking, moving things etc. An industrial robots is best at repetitive tasks, tasks that require precision movements, moving heavy loads and working in hazardous conditions.

1.5 Exercise
(1) Briefly explain, the various components of artificial intelligence (2) Explain the different components of expert systems (3) What are the various applications of expert systems (4) Briefly explain the neural network process (5) Explain robotics
Management Information System Block II Page No.: 13

Você também pode gostar