' !

! .%&"

$

Edited by:

[nom]

[nom] TO [nom]

INTELLIGENCE & EXPERT SYSTEMS

[nom]

Dr. Anil Sharma

Printed by

EXCEL BOOKS PRIVATE LIMITED

A-45, Naraina, Phase-I, New Delhi-110028 for Lovely Professional University

Phagwara

SYLLABUS

[nom] to [nom] Intelligence & Expert Systems

Objectives:

• To enable the student to understand technicalities of intelligence.
• To enable the student to understand techniques of capturing and generating knowledge.
• To enable the student to understand knowledge representation methodologies.
• To enable the student to learn Natural language processing.
• To enable the student to learn Fuzzy Logic with their applications.
• To enable the student to understand probabilistic reasoning.
• To enable the student to learn technicalities of expert system.
• To enable the student to understand [nom] intelligence language 'LISP' and 'Prolog'.

Sr. No. · Description

• · Overview of AI: What is AI, Importance of AI, Early Work in AI, AI and Related Fields.
• · Knowledge: General Concepts, [nom], Definition and Importance of Knowledge, Knowledge-based Systems, Representation of Knowledge, Knowledge Organization, Knowledge Manipulation, Acquisition of Knowledge.
• · LISP and Other AI Programming Languages: [nom] to LISP Syntax and Numeric Functions, Basic List Manipulation Functions in LISP, Functions, Predicates, and Conditionals, Input, Outputs and Local Variables, Iterations and Recursion, Property Lists and Arrays, PROLOG and Other AI Programming Languages.
• · Formalized Symbolic Logics: [nom], Syntax and Semantics for Propositional Logic, Syntax and Semantics for FOPL, Properties of WFFs, Conversion to Clausal Form, Inference Rules, The Resolution Principle, Non-deductive Inference Methods, Representations using Rules Dealing with Inconsistencies and Uncertainties:

Truth Maintenance System, Predicated Completion and Circumscription, Modal and Temporal Logics, Fuzzy

Logic and Natural Language Computation.

• · Probabilistic Reasoning: Bayesian Probabilistic Inference, Possible World Representations, Dempster-Shafer Theory, Ad-Hoc Methods, Heuristic Reasoning Methods. Structured Knowledge: Associative Networks, Frame Structures, Conceptual Dependencies and Scripts. Object Oriented Representation: Overview of Object-Oriented Systems, Object, Classes, Messages and Methods.
• · Search and Control Strategies: Preliminary Concepts, Examples of Search Problems, Uninformed or Blind Search, Informed Search, Searching And-Or Graph.
• · Matching Techniques: Structures used in Matching, Measures for Matching, Matching Like Patterns, Partial Matching, Fuzzy Matching Algorithms, The RETE Matching Algorithm. Knowledge Organization and Management: Indexing and Retrieval Techniques, Integrating Knowledge in Memory.
• · Natural Language Processing: Overview of Linguistics, Grammars and Languages, Basic Parsing Techniques, Semantic Analysis and Representation Structures, Natural Language Generalization, Natural Language Systems, Recognition and Classification Process.
• · Expert System Architecture: Rule-Based Architecture, Nonproduction System Architectures, Dealing with Uncertainty, Knowledge Acquisition and Validation.
• · Types of Learning, Knowledge Acquisition is Difficult, General Learning Model, Performance Measures, Knowledge System Building Tools, Learning by Induction: Generalization and Specialization, Inductive Bias.

Analogical Reasoning and Learning, Explanation based Learning.

CONTENT

Unit 1: · Overview of [nom] Intelligence

Anil Sharma, Lovely Professional University

Unit 2: · Knowledge

Anil Sharma, Lovely Professional University

Unit 3: · Representation of Knowledge

Anil Sharma, Lovely Professional University

Unit 4: · LISP

Anil Sharma, Lovely Professional University

Unit 5: · [nom] Intelligence Programming Language

Anil Sharma, Lovely Professional University

Unit 6: · Formalized Symbolic Logics

Mithilesh Kumar Dubey, Lovely Professional University

Unit 7: · Probabilistic Reasoning

Mithilesh Kumar Dubey, Lovely Professional University

Unit 8; · Structured Representation of Knowledge

Dinesh Kumar, Lovely Professional University

Unit 9: · Search and Control Strategies

Anil Sharma, Lovely Professional University

Unit 10: · Matching Techniques

Anil Sharma, Lovely Professional University

Unit 11: · Knowledge Organization and Management

Mithilesh Kumar Dubey, Lovely Professional University

Unit 12: · Natural Language Processing

Mithilesh Kumar Dubey, Lovely Professional University

Unit 13: · Expert System Architecture

Mithilesh Kumar Dubey, Lovely Professional University

Unit 14: · Types of Learning

Anil Sharma, Lovely Professional University

1

17

30

51

71

84

126

147

165

184

200

215

234

253

Anil Sharma, Lovely Professional University · Unit 1: Overview of [nom] Intelligence

Unit 1: Overview of [nom] Intelligence · Notes

CONTENTS

Objectives

[nom]

1.1 · What is AI?

1.1.1 · Goals of AI

1.2 · Importance of [nom] Intelligence

1.2.1 · [nom] Intelligence in Manufacturing

1.2.2 · [nom] Intelligence in Financial Services

1.2.3 · [nom] Intelligence in Marketing

1.2.4 · [nom] Intelligence in HR

1.3 · Early Work in [nom] Intelligence

1.3.1 · Expert Systems

1.3.2 · Natural Language Processing (NLP)

1.3.3 · Computer Vision

1.3.4 · Intelligent Robots

1.3.5 · Industrial Applications

1.3.6 · Computer-aided Instruction (CAI)

1.3.7 · Learning by Computers

1.3.8 · Voice Recognition (VR)

1.4 · AI and Related Fields

1.4.1 · Common Techniques Used in AI

1.4.2 · Related Fields of AI

1.5 · Summary

1.6 · Keywords

1.7 · Review Questions

1.8 · Further Readings

Objectives

After studying this unit, you will be able to:

• Define [nom] Intelligence
• Describe the goals of AI and the importance of [nom] Intelligence
• Elaborate on the Early Work in [nom] Intelligence
• Discuss the Common Techniques Used in AI

LOVELY PROFESSIONAL UNIVERSITY · 1

[nom] to [nom] Intelligence & Expert Systems

Notes · [nom]

[nom] Intelligence (AI) is the intelligence of machines or software and is also a branch of computer science that studies and develops intelligent machines and software. This field of computer science, AI, is defined as “the study and design of intelligent agents” where an intelligent agent is a system that perceives its environment and takes actions that maximize its chances of success. It is also defined as “the science and engineering of making intelligent machines”.

AI research is highly technical and specialized, deeply divided into subfields that often fail to communicate with each other. Some of the division is due to social and cultural factors: subfields have grown up around particular institutions and the work of individual researchers. AI research is also divided by several technical issues. There are subfields which are focused on the solution of specific problems, on one of several possible approaches, on the use of widely differing tools and towards the accomplishment of particular applications. There is no established unifying theory or paradigm that guides AI research. Researchers disagree about many issues A few of the most long standing questions that have remained unanswered are these: should [nom]

intelligence simulate natural intelligence by studying psychology or neurology?

The central problems (or goals) of AI research include reasoning, knowledge, planning, learning, communication, perception and the ability to move and manipulate objects. General intelligence or “strong AI” is still among the field’s long term goals. Currently popular approaches include statistical methods, computational intelligence and traditional symbolic AI. There are an enormous number of tools used in AI, including versions of search and mathematical optimization, logic, methods based on probability and economics, and many others. In this unit, you will be able to understand the meaning of [nom] Intelligence, its goals and importance, the early work and common techniques used in [nom] Intelligence.

1.1 What is AI?

[nom] intelligence (AI) is the science of programming computers to perform complex tasks that normally require human intelligence. [nom] intelligence systems used in information security applications include [nom] Neural Networks, expert systems and genetic algorithms.

AI is an emerging technology that has recently attracted considerable publicity. Many applications are now under development. One simple view of AI is that it is concerned with devising computer programs to make computers smarter. Thus, research in AI is focused on developing computational approaches to intelligent behavior. This research has two goals:

• · Making machines more useful.
• · Understanding intelligence.

The computer programs with which AI is concerned are primarily symbolic processes involving complexity, uncertainty, and ambiguity. These processes are usually those for which algorithmic solutions do not exist and search is required. Thus, AI deals with the types of problem-solving and decision-making that humans continually face in dealing with the world.

1.1.1 Goals of AI

The general problem of simulating (or creating) intelligence has been broken down into a number of specific sub-problems. These consist of particular traits or capabilities that researchers would like an intelligent system to display. The traits described below have received the most attention.

2 · LOVELY PROFESSIONAL UNIVERSITY

Unit 1: Overview of [nom] Intelligence

Deduction, Reasoning and Problem Solving · Notes

Early AI researchers developed algorithms that imitated the step-by-step reasoning that humans use when they solve puzzles or make logical deductions. By the late 1980s and 1990s, AI research had also developed highly successful methods for dealing with uncertain or incomplete information, employing concepts from probability and economics.

For difficult problems, most of these algorithms can require enormous computational resources – most experience a “combinatorial explosion”: the amount of memory or computer time required becomes astronomical when the problem goes beyond a certain size. The search for more efficient problem-solving algorithms is a high priority for AI research.

Human beings solve most of their problems using fast, intuitive judgments rather than the conscious, step-by-step deduction that early AI research was able to model. AI has made some progress at imitating this kind of “sub-symbolic” problem solving: embodied agent approaches emphasize the importance of sensory motor skills to higher reasoning; neural net research attempts to simulate the structures inside the brain that give rise to this skill; statistical approaches to AI mimic the probabilistic nature of the human ability to guess.

Knowledge Representation

Ontology represents knowledge as a set of concepts within a domain and the relationships between those concepts.

Figure 1.1: Knowledge Representation

Entity

Set · Item

Individual · Category

Abstract · Concrete · Spacetime

relator · Property · Occurrent · Presential

Knowledge representation and knowledge engineering are central to AI research. Many of the problems machines are expected to solve will require extensive knowledge about the world.

Among the things that AI needs to represent are: objects, properties, categories and relations between objects; situations, events, states and time; causes and effects; knowledge about knowledge and many other, less well researched domains. A representation of “what exists” is ontology: the set of objects, relations, concepts and so on that the machine knows about. The most general are called upper ontologism, which attempt provides a foundation for all other knowledge.

LOVELY PROFESSIONAL UNIVERSITY · 3

[nom] to [nom] Intelligence & Expert Systems

Notes · 1.2 Importance of [nom] Intelligence

[nom] Intelligence deals with study and building of rational systems. The [nom] Intelligence has the following advantages:

(i) · This method is more general as compared to the laws of thought method.

(ii) · This method responds to science than to the methods based on human behaviour or

thought.

(iii) · The first computer program introduced to beat the champion chess player was Hitech.

Limited intellectual skills are required for the chess strategies.

(iv) · [nom] Intelligence is divided into many branches. [nom] Intelligence search programs

to study the various probabilities which include chess game moves. Different ways are being set up to carry out more proficiently. Logical [nom] Intelligence deals with mathematical logical program. A program of pattern recognition studies and compares with a particular pattern. It matches the eyes and nose to the face of a police id. Epistemology deals with the study of different types of knowledge required to solve problems.

(v) · Currently, voice recognition program failed to recognize every word but is slowly leading

towards improvement. This idea guide disabled people to utilize computers.

(vi) · Now computers perform many tasks that were beyond our imagination. By 2020, computers

will be more intelligent than the mankind. Due to these advances we will not be inferior to computers as we can turn it off when it fails to perform.

1.2.1 [nom] Intelligence in Manufacturing

AI in manufacturing is wide. It can be described an interactive note-taking system for pen-based computers with two distinctive features. The system is an example of a learning apprentice software-agent. A machine learning component characterises the syntax and semantics of the users information. In machine learning that acquires knowledge for some of the higher level processing; it was found that performance equals that of a partially trainable discourse module requiring manual customisation for each domain. The central assumption of the model is that the learner is embedded within an environment of related learning tasks. Explicit bounds were also derived demonstrating that learning multiple tasks within an environment of related tasks could potentially give much better generalization than learning a [information personnelle] task.

Knowledge and Intelligence are as fundamental as the universe within which they exist, it may turn out that they are more fundamental. Enterprises that utilize AI-enhanced applications are expected to become more diverse, as the needs for the ability to analyze data across multiple variables, fraud detection and customer relationship management emerge as key business drivers to gain competitive advantage. [nom] Intelligence is a branch of Science which deals with helping machines, finds solutions to complex problems in a more human-like fashion. This generally involves borrowing characteristics from human intelligence, and applying them as algorithms in a computer-friendly way. A more or less flexible or efficient approach can be taken depending on the requirements established.

1.2.2 [nom] Intelligence in Financial Services

Banks use [nom] intelligence systems to organize operations, invest in stocks and manage properties. In August 2001, robots beat humans in a simulated financial trading competition.

Financial institutions have long used [nom] neural network systems to detect charges or claims outside of the norm, flagging these for human investigation.

4 · LOVELY PROFESSIONAL UNIVERSITY

Unit 1: Overview of [nom] Intelligence

Creative Virtual has deployed [nom] intelligence customer support systems, or automated · Notes

online assistants at HSBC and Lloyds Banking Group, to assist financial services customers with services such as checking an account balance, signing up for a new credit card or retrieving a forgotten password.

1.2.3 [nom] Intelligence in Marketing

[nom] intelligence is a field of study that “seeks to explain and emulate intelligent behavior in terms of computational processes” through performing the tasks of decision making, problem solving and learning. Unlike other fields associated with intelligence, [nom] intelligence is concerned with both understanding and building of intelligent entities, and has the ability to automate intelligent processes. It is evident that [nom] intelligence is impacting on a variety of subfields and wider society. However, literature regarding its application to the field of marketing appears to be scarce. Advancements in [nom] intelligence’s application to a range of disciplines have led to the development of [nom] intelligence systems which have proved useful to marketers. These systems assist in areas such as market forecasting, automation of processes and decision making and increase the efficiency of tasks which would usually be performed by humans. The science behind these systems can be explained through neural networks and expert systems which are computer programs that process input and provide valuable output for marketers. In the area of social networking, AI is used to [nom] intelligence systems stemming from Social computing technology can be applied to understand social networks on the Web. Data mining techniques can be used to analyze different types of social networks. This analysis helps a marketer to identify influential actors or nodes within networks, this information can then be applied to take a Societal marketing approach. [nom] intelligence has gained significant recognition in the marketing industry. However, ethical issues surrounding these systems and their potential to impact on the need for humans in the workforce, specifically marketing, is a controversial topic. AI-enhanced analytics programs also provide survival modeling capabilities suggesting changes to products based on use.

Example: Customer patterns are analyzed to learn ways to extend the life of light bulbs or to help decide the correct dosage for medications. High-tech data mining can give companies a precise view of how particular segments of the customer base react to a product or service and propose changes consistent with those findings.

1.2.4 [nom] Intelligence in HR

[nom] Intelligence (AI) has been used in business applications since the early eighties. As with all technologies, AI initially generated much interest, but failed to live up to the hype.

However, with the advent of web-enabled infrastructure and rapid strides made by the AI development community, the application of AI techniques in real-time business applications has picked up substantially in the recent past.

Computers are fundamentally well suited to performing mechanical computations, using fixed programmed rules. This allows [nom] machines to perform simple monotonous tasks efficiently and reliably, which humans are ill-suited to. For more complex problems, things get more difficult. Unlike humans, computers have trouble understanding specific situations, and adapting to new situations. [nom] Intelligence aims to improve machine behavior in tackling such complex tasks.

Together with this, much of AI research is allowing us to understand our intelligent behavior.

Humans have an interesting approach to problem-solving, based on abstract thought, high-level deliberative reasoning and pattern recognition. [nom] Intelligence can help us understand

LOVELY PROFESSIONAL UNIVERSITY · 5

[nom] to [nom] Intelligence & Expert Systems

Notes · this process by recreating it, then potentially enabling us to enhance it beyond our current

capabilities. The main purposes of the study are to discuss the appointment of managers in enterprises through fuzzy neural network, to construct a new model for evaluation of managerial talent, and accordingly to develop a decision support system in human resource selection.

Therefore, the research methods of reviewing literature, in-depth interview, questionnaire survey, and fuzzy neural network are used in the study.

Task · Go through various websites for the definition of AI.

!

Caution Have proper knowledge of all procedure before applying the AI.

Self Assessment

State whether the following statements are true or false:

• · AI can be defined as “the science and engineering of making intelligent machines”.
• · Knowledge representation and knowledge engineering are central to AI research.
• · [nom] Intelligence has no branch.
• · Banks use [nom] intelligence systems to organize operations, invest in stocks and manage properties.
• · AI deals with the types of problem-solving and decision-making that humans continually face in dealing with the world.

1.3 Early Work in [nom] Intelligence

1.3.1 Expert Systems

A hierarchical control system is a form of control system in which a set of devices and governing software is arranged in a hierarchy. Intelligent agents must be able to set goals and achieve them. They need a way to visualize the future (they must have a representation of the state of the world and be able to make predictions about how their actions will change it) and be able to make choices that maximize the utility or “value” of the available choices.

Figure 1.2: A Planned System

Hierarchical Control System top level node tasks, goals

sensations,

results

node · node

sensor · actuator · sensor/actuator

actions · actions

sensations · sensations

Controlled system, controlled process, or environment

6 · LOVELY PROFESSIONAL UNIVERSITY

Unit 1: Overview of [nom] Intelligence

In classical planning problems, the agent can assume that it is the only thing acting on the world · Notes

and it can be certain what the consequences of its actions may be. However, if the agent is not the only actor, it must periodically ascertain whether the world matches its predictions and it must change its plan as this becomes necessary, requiring the agent to reason under uncertainty.

1.3.2 Natural Language Processing (NLP)

Natural Language Processing gives machines the ability to read and understand the languages that humans speak. A sufficiently powerful natural language processing system would enable natural language user interfaces and the acquisition of knowledge directly from human-written sources, such as Internet texts.

Figure 1.3: A NLP System

A parse tree represents the syntactic structure of a sentence according to some formal grammar.

Some straightforward applications of natural language processing include information retrieval

(or text mining) and machine translation.

A common method of processing and extracting meaning from natural language is through semantic indexing. Increases in processing speeds and the drop in the cost of data storage makes indexing large volumes of abstractions of the users input much more efficient.

1.3.3 Computer Vision

Machine learning is the study of computer algorithms that improve automatically through experience and has been central to AI research since the field’s inception. Unsupervised learning is the ability to find patterns in a stream of input. Supervised learning includes both classification and numerical regression. Classification is used to determine what category something belongs in, after seeing a number of examples of things from several categories. Regression is the attempt to produce a function that describes the relationship between inputs and outputs and predicts how the outputs should change as the inputs change. In reinforcement learning, the agent is rewarded for good responses and punished for bad ones. These can be analyzed in terms of decision theory, using concepts like utility. The mathematical analysis of machine learning algorithms and their performance is a branch of theoretical computer science known as computational learning theory.

1.3.4 Intelligent Robots

[nom] Intelligence (AI) is a general term that implies the use of a computer to model and/or replicate intelligent behavior. Research in AI focuses on the development and analysis of algorithms that learn and/or perform intelligent behavior with minimal human intervention.

LOVELY PROFESSIONAL UNIVERSITY · 7

[nom] to [nom] Intelligence & Expert Systems

Notes · These techniques have been and continue to be applied to a broad range of problems that arise

in robotics, e-commerce, medical diagnosis, gaming, mathematics, and military planning and logistics, to name a few.

Several research groups fall under the general umbrella of AI in the department, but are disciplines in their own right, including: robotics, NLP, computer vision, computational biology, and e-commerce. Specifically, research is being conducted in estimation theory, mobility mechanisms, multi-agent negotiation, natural language interfaces, machine learning, active computer vision, probabilistic language models for use in spoken language interfaces, and the modeling and integration of visual, haptic, auditory and motor information.

1.3.5 Industrial Applications

[nom] intelligence is implemented in automated online assistants that can be seen as avatars on web pages. It can avail for enterprises to reduce their operation and training cost. A major underlying technology to such systems is natural language processing. Similar techniques may be used in answering machines of call centers, such as speech recognition software to allow computers to handle first level of customer support, text mining and natural language processing to allow better customer handling, agent training by automatic mining of best practices from past interactions, support automation and many other technologies to improve agent productivity and customer satisfaction. Fuzzy logic controllers have been developed for automatic gearboxes in automobiles. Many telecommunications companies make use of heuristic search in the management of their workforces, for example BT Group has deployed heuristic search in a scheduling application that provides the work schedules of 20,000 engineers. The 1990s saw some of the first attempts to mass-produce domestically aimed types of basic [nom] Intelligence for education or leisure. This prospered greatly with the Digital Revolution and helped introduce people, especially children, to a life of dealing with various types of [nom] Intelligence, specifically in the form of Internet. The evolution of music has always been affected by technology.

1.3.6 Computer-aided Instruction (CAI)

With AI, scientists are trying to make the computer emulate the activities of the skillful musician.

Composition, performance, music theory, sound processing are some of the major areas on which research in Music and [nom] Intelligence are focusing. The Air Operations Division uses AI for the rule based expert systems. The AOD has use for [nom] intelligence for surrogate operators for combat and training simulators, mission management aids, support systems for tactical decision making, and post processing of the simulator data into symbolic summaries.

To truly learn is to digest and make the material one’s own by updating one’s internal models and using them in new applications. Real-time interaction with a computer providing immediate feedback and individual guidance is particularly appropriate to this goal. Thus, as computer hardware costs continue to tumble, the nature of the entire present educational system may be radically changed.

1.3.7 Learning by Computers

Computers can support the variety of ways learners construct their own understanding. Students who gather information from the Internet can be self-directed and independent. They can choose what sources to examine and what connections to pursue. Depending on the parameters set by teachers, the students may be in complete control of their topics and their explorations.

Students can work through a computer-based activity at their own pace. Rather than 25 individuals working together on one activity, technology allows independent completion of work. Those

8 · LOVELY PROFESSIONAL UNIVERSITY

Unit 1: Overview of [nom] Intelligence

who begin to fall behind can receive an instructor’s individualized attention while others can · Notes

begin to tackle more complex tasks.

Computer software can mix text, pictures, sound, and motion to provide a variety of options for learners. Multimedia software will not be the only classroom resource, but it can contribute richness and variety to student work.

Students can build on their own understanding by using computers as resource tools, as work stations for individual learning, or as communication channels to share their ideas with other learners. Individual understanding and experiences must be shared and compared to curriculum content. By uncovering students’ individual understandings, teachers can determine the influence of students’ prior knowledge and further their education through new experience.

Computers can be used to assist active experiences – gathering data and resources, conversing with colleagues, struggling through a challenging puzzle or application – or they can assist in reflection. For example, while an online conversation through e-mail is an active event, such discussions usually prompt reflection. They help us think about ideas and check our understanding. In another reflective application, teachers can enlist computers as authoring tools for students’ journals which are excellent vehicles for thoughtful examination of experience.

Notes Introducing technology into the learning environment can encourage cooperative learning and student collaboration. If they are allowed to converse, most students like to talk about their computer work and share their strategies. Classroom activities that are structured so that computers encourage collaboration build on learners’ desire to communicate and share their understanding. It takes planning and intervention to build successful cooperative groups with or without computers, but groups that use computers as teamwork tools have a better start toward collaborative work.

Beyond the classroom, computer networking allows students to communicate and collaborate with content experts and with fellow students around the globe. Communication tools like e-mail, bulletin boards, and chat groups allow teachers to exchange lesson plans and teaching strategies and create a professional community.

The use of real world tools, relevant experiences, and meaningful data inject a sense of purpose to classroom activity. Part of the mission of educational institutions is to produce workforce-ready graduates who can, among other things, manipulate and analyze raw data, critically evaluate information, and operate hardware and software. This technological literacy imparts a very important set of vocational skills that will serve students well in the working world.

Technology has allowed schools to provide greater assistance to traditionally underserved populations. Assistive technology such as voice recognition systems, dynamic Braille displays, speech synthesizers, and talking books provide learning and communication alternatives for those who have developmental or physical disabilities. Research has also shown that computer-mediated communication can ease the social isolation that may be experienced by those with disabilities. Computers have proved successful in increasing academic motivation and lessening anxiety among low ability students and learning disabled students, many of whom simply learn in a manner different from that practiced in a traditional, non-technological classroom.

1.3.8 Voice Recognition (VR)

In computer science, voice recognition is the translation of spoken words into text. It is also known as “automatic speech recognition”, “ASR”, “computer speech recognition”, “speech to

LOVELY PROFESSIONAL UNIVERSITY · 9

[nom] to [nom] Intelligence & Expert Systems

Notes · text”, or just “STT”. Some SR systems use “training” where an individual speaker reads sections

of text into the SR system. These systems analyze the person’s specific voice and use it to fine tune the recognition of that person’s speech, resulting in more accurate transcription. Systems that do not use training are called “Speaker Independent” systems. Systems that use training are called “Speaker Dependent” systems. Speech recognition applications include voice user interfaces such as voice dialing (e.g. “Call home”), call routing (e.g. “I would like to make a collect call”), demotic appliance control, search (e.g. find a podcast where particular words were spoken), simple data entry (e.g., entering a credit card number), preparation of structured documents

(e.g. a radiology report), speech-to-text processing (e.g., word processors or emails), and aircraft

(usually termed Direct Voice Input). The term “voice recognition” refers to finding the identity of “who” is speaking, rather than what they are saying. Recognizing the speaker can simplify the task of translating speech in systems that have been trained on specific person’s voices or it can be used to authenticate or verify the identity of a speaker as part of a security process.

Self Assessment

State whether the following statements are true or false:

• · Natural Language Processing (NLP) gives machines the ability to read and understand the languages that humans speak.
• · Unsupervised learning includes both classification and numerical regression.
• · The Air Operations Division uses AI for the rule based expert systems.
• · Voice Recognition (VR) is also known as “automatic speech recognition” or “speech to text”.

1.4 AI and Related Fields

The AI may be defined as a branch of Computer Science that is concerned with automation of intelligent behavior. Its principles include the data structures used and knowledge representation, the algorithms needed to apply the knowledge, and language and programming techniques used in their implementation.

1.4.1 Common Techniques Used in AI

Neural Networks: Neural networks are structures that can be “trained” to recognize patterns in inputs. Neural networks typically take a vector of input values and produce a vector of output values. Inside, they train weights of “neurons”. Neural networks use supervised learning, in which inputs and outputs are known and the goal is to build a representation of a function that will approximate the input to output mapping.

They are a way to implement function approximation: given y 1 = f(x 1 ), y 2 = f(x 2 ), …, y n = f(x n ), construct a function f’ that approximates f. The approximate function f’ is typically smooth: for x’ close to x, we will expect that f’(x’) is close to f’(x). Function approximation serves two purposes:

• · Size: The representation of the approximate function can be significantly smaller than the true function.
• · Generalization: The approximate function can be used on inputs for which we do not know the value of the function.

In path finding, the function is f(start, goal) = path. We do not already know the output paths. We could compute them in some way, perhaps by using A*. But if we are able to compute a path given (start, goal), then we already know the function f, so why bother approximating it?

10 · LOVELY PROFESSIONAL UNIVERSITY

Unit 1: Overview of [nom] Intelligence

There is no use in generalizing f because we know it completely. The only potential benefit · Notes

would be in reducing the size of the representation of f. The representation of f is a fairly simple algorithm, which takes little space, so I don’t think that’s useful either. In addition, neural networks produce a fixed-size output, whereas paths are variable sized.

Genetic Algorithms: Genetic Algorithms allow to explore a space of parameters to find solutions that score well according to a “fitness function”. They are a way to implement function optimization: given a function g(x) (where x is typically a vector of parameter values), find the value of x that maximizes (or minimizes) g(x). This is an unsupervised learning problem. The right answer is not known beforehand. For path finding, given a starting position and a goal, x is the path between the two and g(x) is the cost of that path. Simple optimization approaches like hill-climbing will change x in ways that increase g(x). Unfortunately, in some problems, we reach “local maxima”, values of x for which no nearby x has a greater value of g, but some faraway value of x is better. Genetic algorithms improve upon hill-climbing by maintaining multiple x, and using evolution-inspired approaches like mutation and crossover to alter x. Both hill-climbing and genetic algorithms can be used to learn the best value of x. For path finding, however, we already have an algorithm (A*) to find the best x, so function optimization approaches are not needed.

Genetic Programming takes genetic algorithms a step further, and treats programs as the parameters. For example, we would breed path finding algorithms instead of paths, and your fitness function would rate each algorithm based on how well it does. For path finding, we already have a good algorithm and we do not need to evolve a new one.

It may be that as with neural networks, genetic algorithms can be applied to some portion of the path finding problem.

Reinforcement Learning: Like genetic algorithms, Reinforcement Learning is an unsupervised learning problem. However, unlike genetic algorithms, agents can learn during their lifetimes;

it’s not necessary to wait to see if they “live” or “die”. Also, it’s possible for multiple agents experiencing different things to share what they’ve learned. Reinforcement learning has some similarities to the core of A*. In A*, reaching the end goal is propagated back to mark all the choices that were made along the path; other choices are discarded. In reinforcement learning, every state can be evaluated and its reward (or punishment) is propagated back to mark all the choices that were made leading up to that state. The propagation is made using a value function, which is somewhat like the heuristic function in A*, except that it’s updated as the agents try new things and learn what works. One of the key advantages of reinforcement learning and genetic algorithms over simpler approaches is that there is a choice made between exploring new things and exploiting the information learned so far. In genetic algorithms, the exploration via mutation; in reinforcement learning, the exploration is via explicitly allowing the probability of choosing new actions. As with genetic algorithms, we don’t believe reinforcement learning should be used for the path finding problem itself, but instead as a guide for teaching agents how to behave in the game world.

1.4.2 Related Fields of AI

Following may be the related fields of AI.

(a) · Robotics

(b) · Computer Vision

(c) · Image Processing

(d) · Voice Recognition

(e) · Neural Networks

LOVELY PROFESSIONAL UNIVERSITY · 11

[nom] to [nom] Intelligence & Expert Systems

Notes · (a) · [nom] Intelligence – Robotics: Robots are a practical application of [nom] intelligence

of rapidly growing importance for industrial, domestic, entertainment and military tasks.

The capabilities of today’s robots go far beyond the factory robots of the last century, whose operations depended on the strictly controlled conditions of an assembly line.

Today’s field and service robots must be able operate in the unstructured world of a shop floor, a home, a school or a battlefield. Furthermore, the software controlling such machines must be far more flexible, robust and autonomous than the robot programming languages of the past. They must be able to be commanded to go about their tasks far more readily than is today possible. In order to solve such demanding problems, it is not only necessary to build and test new kinds machines, but also to create new and powerful classes of algorithms, the general applicability of which can be demonstrated by the successful operation of those machines on challenging problems. We invite research students to join us in this important work; skills in electronics, mechanics or software development are welcome, but an interest in robots is essential!

(b) · [nom] Intelligence – Vision: Computer vision is a field that includes methods for

acquiring, processing, analyzing, and understanding images and, in general, high-dimensional data from the real world in order to produce numerical or symbolic information, e.g., in the forms of decisions. A theme in the development of this field has been to duplicate the abilities of human vision by electronically perceiving and understanding an image. This image understanding can be seen as the disentangling of symbolic information from image data using models constructed with the aid of geometry, physics, statistics, and learning theory. Computer vision has also been described as the enterprise of automating and integrating a wide range of processes and representations for vision perception.

Applications range from tasks such as industrial machine vision systems which, say, inspect bottles speeding by on a production line, to research into [nom] intelligence and computers or robots that can comprehend the world around them. The computer vision and machine vision fields have significant overlap. Computer vision covers the core technology of automated image analysis which is used in many fields. Machine vision usually refers to a process of combining automated image analysis with other methods and technologies to provide automated inspection and robot guidance in industrial applications.

As a scientific discipline, computer vision is concerned with the theory behind [nom]

systems that extract information from images. The image data can take many forms, such as video sequences, views from multiple cameras, or multi-dimensional data from a medical scanner.

As a technological discipline, computer vision seeks to apply its theories and models to the construction of computer vision systems. Examples of applications of computer vision include systems for:

• Controlling processes, e.g., an industrial robot;
• Navigation, e.g., by an autonomous vehicle or mobile robot;
• Detecting events, e.g., for visual surveillance or people counting;
• Organizing information, e.g., for indexing databases of images and image sequences;
• Modeling objects or environments, e.g., medical image analysis or topographical modeling;
• Interaction, e.g., as the input to a device for computer-human interaction, and
• Automatic inspection, e.g., in manufacturing applications.

12 · LOVELY PROFESSIONAL UNIVERSITY

Unit 1: Overview of [nom] Intelligence

(c) · [nom] Intelligence – Image Processing: We can apply [nom] intelligence techniques · Notes

in digital image fundamentals, image enhancement, image restoration, morphological image processing, image segmentation and edge detection, object recognition, image representation and description, colour image processing, wavelets and multi resolution processing, image compression.

(d) · [nom] Intelligence – Voice Recognition: Voice recognition has existed in some form or

another since the 1950s, Since then, many other companies have toyed with voice recognition, including Dragon Dictation, which launched the first speech recognition software for the PC. Soon after, the telecom industry began creating voice portals, which were intended to replace customer service representatives by giving information through voice-activated menus. Instead, they became a nuisance to cell phone users looking for answers about their overage charges.

Despite these advances, the ability to truly communicate with machines through conversation remained confined to the realm of science fiction.

The latest advances in voice recognition software have found a niche in the automotive department. Ford’s SYNC feature debuted in 2007 and has now become a popular offering across its product line. In last year’s Super Bowl, Chevrolet advertised On Star’s ability to read live Facebook feeds aloud.

AI for speech recognition involves two basic ideas. Firstly, it involves studying the thought processes of human beings. Secondly, it deals with representing those processes via machines (like computers, robots, etc.). AI is behavior of a machine, which, if performed by a human being, would be called intelligence. It makes machines smarter and more useful, and is less expensive than natural intelligence. NLP, refers to [nom] intelligence methods of communicating with a computer in a natural language like English. The main objective of a NLP program is to understand input and initiate action. The input words are scanned and matched against internally stored known words. Identification of a keyword causes some action to be taken. In this way, one can communicate with the computer in one’s language.

(e) · [nom] Intelligence – Neural Network: A neural network is, in essence, an attempt to

simulate the brain. Neural network theory revolves around the idea that certain key properties of biological neurons can be extracted and applied to simulations, thus creating a simulated (and very much simplified) brain. The first important thing to understand then is that the components of an [nom] neural network are an attempt to recreate the computing potential of the brain. The second important thing to understand, however, is that no one has ever claimed to simulate anything as complex as an actual brain. Whereas the human brain is estimated to have something on the order of ten to a hundred billion neurons, a typical [nom] Neural Network (ANN) is not likely to have more than 1,000

[nom] neurons. While many types of [nom] neural nets exist, most are organized according to the same basic structure. There are three components to this organization: a set of input nodes, one or more layers of ‘hidden’ nodes, and a set of output nodes. The input nodes take in information, and are akin to sensory organs. Whether the information is in the form of a digitized picture, or a series of stock values, or just about any other form that can be numerically expressed, this is where the net gets its initial data. The information is supplied as activation values, that is, each node is given a number, higher numbers representing greater activation. This is just like human neurons except that rather than conveying their activation level by firing more frequently, as biological neurons do,

[nom] neurons indicate activation by passing this activation value to connected nodes.

After receiving this initial activation, information is then passed through the network.

LOVELY PROFESSIONAL UNIVERSITY · 13

[nom] to [nom] Intelligence & Expert Systems

Notes

Did u know? The philosophy of [nom] intelligence can determine the following:

• A machine can act intelligently? Can it solve any problem that a person would solve by thinking?
• A human intelligence and machine intelligence are the same? Is the human brain essentially a computer?
• A machine has a mind, mental states and consciousness in the same sense humans can do? Can it feel how things are?
• The art of creating machines that perform functions that require intelligence when performed by people’’ (Kurzweil, 1990)

Task · Create a list of software that can solve the problems intelligently.

Self Assessment

State whether the following statements are true or false:

• · Neural networks do not use supervised learning.
• · Genetic Programming takes genetic algorithms and treats programs as the parameters.
• · Robots are one of the applications of AI.
• · Human brain is estimated to have something on the order of hundred billion neurons called [nom] neural network (ANN).

Lab Exercise · 1. · Create a list of five existing expert systems and their work domain.

• · Draw a diagram showing knowledge in a system and the place of AI.

Case Study · Application Areas of AI Technology

[nom] intelligence is an integrated part of our daily life and of many fields in research. In archaeology, however, it does not (yet) play an important role. In the past twenty years archaeologists have discussed the potentials of, in particular, expert systems. They have developed some valuable systems, but the general impression is that archaeology is not a suitable host discipline for knowledge-based approaches. In Archaeology and the Application of [nom] Intelligence: case studies on use-wear analysis of prehistoric flint tools, Dr. M. H. van den Dries sets out to validate this negative conclusion.

She states that since most archaeological applications were mere prototypes and have never been subjected to objective tests, there is hardly any ground for this rather radical inference. In order to ground her conclusion objectively, Van den Dries has built two applications, an expert system and a neural network. She used use-wear analysis of prehistoric tools as the application area. The main objective of the project was to develop

Contd...

14 · LOVELY PROFESSIONAL UNIVERSITY

Unit 1: Overview of [nom] Intelligence

a practical training tool for students. As Van den Dries’ aim was to demonstrate the · Notes

practical applicability of both applications, they were exposed to two objective tests in which replicated stone tools as well as prehistoric artefacts were involved. In one test both experience use-wear analysts and students participated. The outcome of this trail has been compared with the results of all other blind tests that human use-wear analysts have carried out hitherto. An important conclusion of Van den Dries’ study is that both applications perform well, but that the expert system is better equipped for educational tasks, while the neural network approach mainly suits research purposes. Therefore, the expert system application, called WAVES, has been made operational. It already supports students of several archaeology departments around the world in learning use-wear analysis. Based on her findings Van den en Dries subsequently advises archaeologists to exploit better the merits that [nom] intelligence offers them, because it is a means to record unique and valuable expert knowledge, to obtain objective analysis results and to democratize archaeological knowledge.

Questions:

• · What are the application areas of AI ?
• · What is WAVES ?

1.5 Summary

• AI is a branch of science which deals with helping machines find solution to complex solutions.
• AI is generally associated with computer science but it has many important links with other fields such as Mathematics, Psychology and Biology etc.
• It is more general than the “laws of thought” approach, because correct inference is only a useful mechanism for achieving rationality, and not a necessary one. Second, it is more amenable to scientific development than approaches based on human behavior or human thought, because the standard of rationality is clearly defined and completely general.

1.6 Keywords

[nom] Intelligence: [nom] Intelligence is one of the newest disciplines. It was formally initiated in 1956, when the name was coined, although at that point work had been under way for about five years.

Expert System: An expert system is a computer system that emulates the decision-making ability of a human expert. Expert systems are designed to solve complex problems by reasoning about knowledge, like an expert, and not by following the procedure of a developer as is the case in conventional programming.

Natural Language: A natural language (or ordinary language) is any language which arises in an unpremeditated fashion as the result of the innate facility for language possessed by the human intellect. A natural language is typically used for communication, and may be spoken, signed, or written.

1.7 Review Questions

• · Explain [nom] intelligence with an example.
• · What are the advantages of AI?
• · How does AI help in solving complex problems of computer science?

LOVELY PROFESSIONAL UNIVERSITY · 15

[nom] to [nom] Intelligence & Expert Systems

Notes · 4. · Discuss the role of AI in police interrogation.

• · Explain the working of any expert systems.
• · What are the different techniques used in AI?
• · Can AI software help in determining human behavior?

Answers: Self Assessment

• · True · 2. · True
• · False · 4. · True
• · True · 6. · True
• · False · 8. · True
• · True · 10. · False
• · True · 12. · True
• · true

1.8 Further Readings

Books · Deshpande, Neeta (2009), [nom] Intelligence, Technical Publications.

Harris, Michael C. (2010), [nom] Intelligence, Marshall Cavendish.

Rich, Elaine (2004), [nom] Intelligence 3E (Sie). Tata McGraw-Hill Education.

Russell, Stuart (2003), [nom] Intelligence: A Modern Approach, 2/E, Pearson

Education India.

Whitby, Blay (2009), [nom] Intelligence, The Rosen Publishing Group.

Online links · aima.cs.berkeley.edu/

[lien]

[lien]

[lien]

[lien]

16 · LOVELY PROFESSIONAL UNIVERSITY

Anil Sharma, Lovely Professional University · Unit 2: Knowledge

Unit 2: Knowledge · Notes

CONTENTS

Objectives

[nom]

2.1 · General Concepts of Knowledge

2.1.1 · Knowledge Progression

2.1.2 · Knowledge Model

2.2 · Definition and Importance of Knowledge

2.2.1 · Three Kinds of Knowledge

2.2.2 · Uses of Knowledge

2.2.3 · Knowledge-base

2.3 · Knowledge-based Systems (KBS)

2.3.1 · Characteristics of KBS

2.3.2 · Structure of a Knowledge-based System

2.3.3 · KBS Development

2.3.4 · Testing

2.3.5 · Why Build a Knowledge-based System?

2.4 · Summary

2.5 · Keywords

2.6 · Review Questions

2.7 · Further Readings

Objectives

After studying this unit, you will be able to:

• Discuss the general concepts of Knowledge
• Explain the definition and importance of Knowledge
• Describe the concept of Knowledge-based Systems (KBS)

[nom]

In the previous unit, we dealt with the meaning of [nom] intelligence, its goals and importance, the early work and common techniques used in [nom] Intelligence. Knowledge is a familiarity with someone or something, which can include facts, information, descriptions or skills acquired through experience or education. It can refer to the theoretical or practical understanding of a subject. It can be implicit (as with practical skill or expertise) or explicit (as with the theoretical understanding of a subject); it can be more or less formal or systematic. In philosophy, the study of knowledge is called epistemology; the philosopher Plato famously defined knowledge as

LOVELY PROFESSIONAL UNIVERSITY · 17

[nom] to [nom] Intelligence & Expert Systems

Notes · “justified true belief.” However, no [information personnelle] agreed upon definition of knowledge exists, though

there are numerous theories to explain it.

The following quote from Bertrand Russell’s “Theory of Knowledge” illustrates the difficulty in defining knowledge: “The question how knowledge should be defined is perhaps the most important and difficult of the three with which we shall deal. This may seem surprising: at first sight it might be thought that knowledge might be defined as belief which is in agreement with the facts. The trouble is that no one knows what a belief is, no one knows what a fact is, and no one knows what sort of agreement between them would make a belief true. Let us begin with belief.” The definition of knowledge is a matter of ongoing debate among philosophers in the field of epistemology. The classical definition, described but not ultimately endorsed by Plato specifies that a statement must meet three criteria in order to be considered knowledge: it must be justified, true, and believed. Some claim that these conditions are not sufficient, as Gettier case examples allegedly demonstrate.

There are a number of alternatives proposed, including Robert Nozick’s arguments for a requirement that knowledge ‘tracks the truth’ and Simon Blackburn’s additional requirement that we do not want to say that those who meet any of these conditions ‘through a defect, flaw, or failure’ have knowledge. Richard Kirkham suggests that our definition of knowledge requires that the evidence for the belief necessitates its truth. In this unit, you will be able to understand the concepts of knowledge progression and model, importance of knowledge, characteristics and structure of Characteristics of KBS.

2.1 General Concepts of Knowledge

The theory of knowledge and creativity is an important department of philosophy. It arose historically with philosophy, as its core, around which everything else was built. This department of philosophy considers a wide range of problems: the relationship between knowledge and reality, its sources and driving forces, its forms and levels, the principles and laws of cognitive activity, and the trends of its development. Philosophy analyses the criteria of the authenticity of knowledge, its veracity, and also the causes of error, the problems of the practical application of knowledge.

As selective reflection of the world cognition expresses the highest creative aspirations of human reason and constitutes the crown of human achievement. Throughout the millennia of its development humanity has travelled a long road, from the primitive and limited to an increasingly profound and comprehensive understanding of the essence of existence.

!

Caution This difficult path has led us to the discovery of innumerable facts, properties and laws of nature, of social life and man himself, to the building of an extremely complex and almost unencompassable scientific picture of the world, to the highly sophisticated sphere of art, to the achievements of modern technology.

Humanity has always striven to acquire new knowledge. The process of mastering the secrets of existence continues unceasingly and its vector is oriented on the infinite vistas of the future. The pace and scale of cognitive activity are constantly increasing. Every day is marked by intellectual advances in a constant quest, which ever more widely and vividly illuminates the remote horizons of the as yet invisible. We are deluged with new discoveries.

The path travelled by science convinces us that the possibilities of human cognition are limitless.

Our reason perceives the laws of the universe in order to bring them under man’s control, in

18 · LOVELY PROFESSIONAL UNIVERSITY

Unit 2: Knowledge

order to refashion the world in the interests of man and society. Human knowledge is a highly · Notes

complex system, a social memory whose wealth is passed on from generation to generation by means of social heredity.

INFERENCE · Control System

ENGINE · Interpreter

KNOWLEDGE

BASE

2.1.1 Knowledge Progression

The concept can be understood by the following figure:

Figure 2.1: Knowledge Progression

Information · Knowledge · Wisdom

Data: Data is viewed as collection of disconnected facts.

Example: It is raining.

Information emerges when relationships among facts are established and understood;

Provides answers to “who”, “what”, “where”, and “when”.

Example: The temperature dropped 15 degrees and then it started raining.

Knowledge emerges when relationships among patterns are identified and understood;

Provides answers as “how”.

Example: If the humidity is very high and the temperature drops substantially, then the atmosphere is unlikely to hold the moisture, so it rains.

Wisdom is the pinnacle of understanding, uncovers the principles of relationships that describe patterns;