l1 kbs csci3406 introduction to kbs l11

19
03/05/2012 1 IMAT3406 Fuzzy Logic and Knowledge Based Systems (AI) Introduction to Knowledge Based Systems (KBS) Most of the KBS notes kindly provided by Dr. Aladdin Ayesh 

Upload: arunnurak

Post on 06-Apr-2018

242 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 1/19

03/05/2012 1

IMAT3406Fuzzy Logic and

Knowledge Based Systems (AI)

Introduction to Knowledge

Based Systems (KBS)

Most of the KBS notes kindly provided by Dr. Aladdin Ayesh 

Page 2: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 2/19

03/05/2012 2

Week 15  Introduction to Knowledge Based Systems (KBS) 

Week 16 – 1  Knowledge Representation I 

Week 16 - 2  Knowledge Representation II 

Week 17 – 1

Week 17 – 2 

Inferencing I (Backward Chaining)(KBS Assignment released and to be submitted by

30th of March 2010 ) (KBS coursework : 10%)

Week 18 – 1 Inferencing II (Forward Chaining) 

Week 18 – 2  Knowledge Acquisition I

Week 19 – 1 Knowledge Acquisition II 

Week 19 – 2 Developing KBS (tools and practical issues)

Week 20 – 1 KBS Applications

Week 20 – 2 Modelling Uncertainty 

Week 27 Revision (FL + KBS)

Lecture Plan for Knowledge Based System

Page 3: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 3/19

03/05/2012 3

Reading ListNot compulsory, but complementary

Knowledge Based Systems E. Turban, Expert Systems and Applied  Artificial Intelligence . New York:

Macmillan Publishing Company, 1992.* T. Dean, J. Allen, and Y. Aloimonos, Artificial Intelligence: Theory and Practice : TheBenjamin/Cummings Publishing Company,

Inc., 1995. P. Jackson, Introduction to Expert Systems ,Second Edition ed. UK: Addison WesleyPublishing Company, 1990.

Page 4: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 4/19

03/05/2012 4

Introduction

In this lecture, we cover an introductionto KBS.

We start with identifying the differenttypes of AI: numerical and symbolic.

We look at some search algorithms as

simple AI system.

Page 5: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 5/19

03/05/2012 5

Topics of Discussion

 AI

Simple AI systems

Developing KBS

Some famous KBS

Page 6: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 6/19

03/05/2012 6

 AI

 Artificial Intelligence is the field of computingthat attempts at providing computationalmodels of some human activities, which

researchers consider intelligent activities,such as learning, acting, decision making,evolving and so on. AI, therefore, relatesstrongly to fields such as psychology, biology

and sociology. In some cases new disciplinesemerged such as bio-informatics andcybernetics.

Page 7: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 7/19

03/05/2012 7

 AI

There are two main streams in developing AIsystems: quantitive and qualitative

approaches.Quantitive approaches sometimes referred toas numerical approaches, because they usequantities in analysing the problems.

Neural nets, fuzzy logic, genetic algorithmsare all examples of the quantitive approach.

Page 8: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 8/19

03/05/2012 8

 AI

Qualitative approaches sometimes referred toas symbolic approaches, because they use

qualities of the problem to solve the problem.Logic, rules, lists based systems are examplesof qualitative AI systems.

Page 9: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 9/19

03/05/2012 9

Simple AI systems

The simplest view of AI systems is as asearch problem solver. It is almost impossible

to develop an expert system withoutimplementing some search technique oranother to navigate through the problemdomain for the solution. Search techniques

provide the base for the inference engine,which is an essential component of anyexpert system.

Page 10: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 10/19

03/05/2012 10

Simple AI systems

There are two main types of searches:Conventional searches and heuristic searches.

Conventional searches cover the entiredomain and eventually find the solution, whatis the problem with that?

Heuristic searches aim at reducing the

domain or covering a selected portion of theproblem domain. What is the problem withthat?

Page 11: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 11/19

03/05/2012 11

Simple AI systems

Conventional searches include:

Depth first search

Breadth first search

Heuristic searches include: Generate and test.

Hill climbing.

Best first.

Problem reduction. Constraint satisfaction.

Means-end analysis.

Page 12: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 12/19

03/05/2012 12

Developing KBS

(Please refer to the second lecture andlecture notes part 2)Many KBS’s are symbolic systems.

There are two distinctive parts need to beincluded in any KBS: Knowledge representation, which is usually

the result of knowledge acquisition

Inference Engine, which you would not usuallyneed to develop if you are using an expertsystem shell such as CLIPS

Page 13: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 13/19

03/05/2012 13

Developing KBS

In KBS, we also call them exact systems, wedo not need to imply certainty factor as wedid in FLS.

In CLIPS, KBS can be developed as pure ruleswithout the need to define fuzzy sets, i.e. nodeftemplate is required.

CLIPS is a productive development and delivery expert system tool whichprovides a complete environment for the construction of rule and/or objectbased expert systems., CLIPS was created in 1985 and is now widely usedthroughout the government, industry, and academia. For further detailsincluding its key features, please see http://www.ghg.net/clips/WhatIsCLIPS.html

Page 14: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 14/19

03/05/2012 14

Some famous KBS

DENDRAL (Late 60s)

MYCIN (Mid 1970s)

R1/XCON (1980s)

Page 15: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 15/19

03/05/2012 15

DENDRAL (1965-83)

DENDRAL (1965-83): The DENDRAL Project was one of the earliest expertsystems. DENDRAL began as an effort to explore the mechanization of 

scientific reasoning and the formalization of scientific knowledge by workingwithin a specific domain of science, organic chemistry. Another concern wasto use AI methodology to understand better some fundamental questions inthe philosophy of science, including the process by which explanatoryhypotheses are discovered or judged adequate. After more than a decade of collaboration among chemists, geneticists, and computer scientists,

DENDRAL had become not only a successful demonstration of the power of rule-based expert systems but also a significant tool for molecular structureanalysis, in use in both academic and industrial research labs. Using a plan-generate-test search paradigm and data from mass spectrometry and othersources, DENDRAL proposes plausible candidate structures for new orunknown chemical compounds. Its performance rivals that of human expertsfor certain classes of organic compounds and has resulted in a number of papers that were published in the chemical literature. Although no longer atopic of academic research, the most recent version of the interactivestructure generator, GENOA, has been licensed by Stanford University forcommercial use.

(taken from http://smi-web.stanford.edu/projects/history.html)

Page 16: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 16/19

03/05/2012 16

MYCIN (1972-80)

MYCIN is an interactive program that diagnoses certain infectious diseases,

prescribes antimicrobial therapy, and can explain its reasoning in detail. In acontrolled test, its performance equalled that of specialists. In addition, theMYCIN program incorporated several important AI developments. MYCINextended the notion that the knowledge base should be separate from theinference engine, and its rule-based inference engine was built on a backward-chaining, or goal-directed, control strategy. Since it was designed as a

consultant for physicians, MYCIN was given the ability to explain both its line of reasoning and its knowledge. Because of the rapid pace of developments inmedicine, the knowledge base was designed for easy augmentation. Andbecause medical diagnosis often involves a degree of uncertainty, MYCIN's rulesincorporated certainty factors to indicate the importance (i.e., likelihood andrisk) of a conclusion. Although MYCIN was never used routinely by physicians, it

has substantially influenced other AI research. At the HPP, MYCIN led to work inTEIRESIAS, EMYCIN, PUFF, CENTAUR, VM, GUIDON, and SACON, all describedbelow, and to ONCOCIN and ROGET. The book Rule-Based Expert Sytem: The MYCIN Experiment at the Stanford Heuristic Programming Project describes thedecade of research on MYCIN and its descendants. (taken from http://smi-web.stanford.edu/projects/history.html)

Page 17: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 17/19

03/05/2012 17

R1/XCON (1980s)

•One of the first commercially successful expert systems• Application domain:

•configuration of minicomputer systems•selection of components

arrangement of components into modules and cases• Approach

•data-driven, forward chaining•consists of about 10,000 rules written in OPS5

•Results

•quality of solutions similar to or better than humanexperts•roughly ten times faster (2 vs. 25 minutes)•estimated savings $25 million/year

Page 18: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 18/19

03/05/2012 18

Conclusion

 AI systems and search algorithms.

Developing KBS.

Page 19: L1 KBS CSCI3406 Introduction to KBS L11

8/3/2019 L1 KBS CSCI3406 Introduction to KBS L11

http://slidepdf.com/reader/full/l1-kbs-csci3406-introduction-to-kbs-l11 19/19

03/05/2012 19

Next Steps

Next … 

Knowledge acquisition.