Knowledge engineering shells : systems and techniques

목차 (Table of Contents)

• CONTENTS
• PROGOGUE = ⅴ
• ANLINTRODUCTION TO KNOWLEDGE ENGINEERING / N. Bourbakis = ⅶ
• CONTRIBUTORS = xi
• SECTION 1 : KNOWLEDGE REPRESENTATION AND MANIPULATION
• Chapter 1 : KNOWLEDGE REPRESENTATION USING SEMANTICS OF SEMANTIC NETWORKS / A. V. Hudli
• 1. Introduction = 3
• 1.1. NVL = 5
• 2. Syntax of NVL = 6
• 2.1. The Basis for NVL Syntax = 6
• 2.2. Symbols Used in NVL = 11
• 2.3. Formation Rules = 11
• 2.4. Axiom schemata and Rules of Inference = 12
• 3. Semantics of NVL Formulas = 15
• 4. Unification of NVL Lists = 17
• 4.1 Parallel Unification = 18
• 4.2 Answer Substitutions = 19
• 5. The NVL Interpreter = 20
• 5.1. Completion of an NVL Program = 20
• 5.2. Simple Inferences = 20
• 5.3. V-Resolution = 21
• 6. Comparing NVL with Predicate Logic = 22
• 6.1. The Relationship between NVL and predicate Logic = 22
• 6.2. Efficiency of NVL = 24
• 7. Increasing the Expressive Power of NVL = 25
• 7.1. Adding Negation = 25
• 7.2. Limitor Lists with Constraints = 26
• 8. Conclusions = 27
• References = 27
• Chapter 2 : AN ONTOLOGICAL APPROACH TO KNOWLEDGE ACQUISITION SCHEMES / I. Monarch and S. Nirenburg
• 1. Requirements for Very Large Dynamic Knowledge Bases = 30
• 2. Categorial and Experienctial Knowledge = 31
• 2.1. Semantic Memory = 31
• 2.2. World Models = 34
• 2.3. Episodic Memory = 36
• 3. Maintenance of VLKBs = 37
• 3.1. Views at Kowledge Representation = 37
• 3.2. The Architecture of a VLKB = 42
• 4. Ac Acquisition of VLKBs = 44
• 4.1. Acquisition Systems and Methodologies = 44
• 4.2. Extensions to ONTOS : Incremental Automation = 45
• References = 49
• Chapter 3 : PRODUCTION RULES AND SYSTEMS : A TOP-DOWN CONSTRUCTION OF BOTTOM-UP INFERENCE / M. Perlin
• 1. Introductioon = 57
• 2. A Concrete Problem = 58
• 3. A Backward Chanining Solution = 60
• 4. A Forward Chaining Solution = 62
• 4.1. Acquiring Production Rules from the Solution Trace = 62
• 4.2. Inductive Learning = 66
• 5. Production Systems= 66
• 6. Rule Interpreters = 68
• 7. The RETE Match Algorithm = 69
• 7.1. Set Filtering = 69
• 7.2. Recursive Set Filtering = 70
• 7.3. Evaluating the Recursion = 72
• 7.4. Network Programs = 73
• 7.5. Incremental Match= 75
• 7.6. Merging Tests = 76
• 8. Extensions to Basic RETE = 77
• 8.1. Control by Data-Driven Agenda = 78
• 8.2. Compiling Tests = 78
• 8.3. Absence Tests = 79
• 8.4. Equality Testing Using Binding Values = 80
• 8.5. Active Tuple Memory = 81
• 9. Conclusions = 82
• References = 83
• Chapter 4: TOWARDS PARALLEL KNOWLEDGE PROCESSING / J.C. Yan
• 1. Introduction = 87
• 1.1. Knowledge Processing Paradigms = 88
• 1.2. Parallel Processing = 90
• 1.3. Chapter Outline = 91
• 2. Sequential Knowledge Processing = 92
• 2.1. Software and Hardware Requirements of Knowledge Processing = 92
• 2.2. LISP Machines=92
• 2.3. PROLOG Machines = 94
• 2.4. The RETE Algorithm= 94
• 2.5. Summary : Sequential Knowledge Processing is Still Too Slow = 95
• 3. Multiprocessors for Knowledge Processing = 96
• 3.1. DADO = 96
• 3.2. NETL = 97
• 3.3. Connection Machine = 97
• 3.4. Parallel Inference Machine = 98
• 3.5. Summary : Parallel Hardware for Knowledge Processing = 99
• 4. Parallel Languages for Knowledge Processing = 99
• 4.1. parallel LISPs = 99
• 4.2. parallel PROLOGs = 101
• 4.3. parallel Object-Oriented Languages = 103
• 4.4. summary : Parallel Languages for Knowledge Processing = 104
• 5. Mapping Knowledge Processing Systems to Multicomputers = 104
• 5.1. Mapping Production Systems onto DADO = 105
• 5.2. Mapping RETE Networks onto Multicomputers = 106
• 5.3. Distributing Rules on Multicomputers = 107
• 5.4. summary : Mapping Expert Systems to Multiprocessors = 108
• 6. Measuring parallelism in Knowledge Processing Systems = 108
• 6.1. Parallelism in Production Systems and Flat Concurrent PROLOG Systems = 109
• 6.2. Obtaining Speed-up via Parallel Processing is Difficult = 110
• 7. Parallel Processing for Knowledge Processing Systems = 111
• 7.1. Speed-up and Parallel Processing = 112
• 7.2. Conclusions = 113
• References = 115
• SECTION 2 : EXPERT SYSTEMS
• Chapter 5 : DEVELOPMENT OF EXPERT SYSTEM SHELLS / J.J.P.Tsai, S.J.H. Yang, N. Waheed and t. Moher
• 1. Introduction = 123
• 2. Knowledge Representation = 124
• 2.1. Rules = 124
• 2.2. Semantic Nets = 125
• 2.3. Frames = 126
• 2.4. Objects = 126
• 3. Inference and Control = 126
• 3.1. Rule Based Control = 127
• 3.2. vp rocedural Control = 127
• 3.3. Object Oriented Control = 128
• 3.4. Access Oriented Control = 128
• 4. Development Environment = 128
• 5. Expert System Shell= 129
• 5.1. Knowledge Engineering Languages = 129
• 5.2. Skeletal Shells = 129
• 5.3. General Purpose Shells = 130
• 5.4. Knowledge Acquisition Tolls = 132
• 6. Trends of Current Expert System Shells = 132
• 7. Description of Some expert System Shells = 133
• 8. List of Commercial Expert System Shells = 141
• References = 150
• Chapter 6 : DESIGNING BETTER METHODOLOGIES FOR EXPERT SYSTEM DEVELOPMENT / W.T.Tsai and I.A. Zualkernan
• 1. Introduction = 153
• 2. Expert System Development = 154
• 3. Essence and Accidental Problems in Expert System Development = 157
• 3.1. Extreme Wicked Problem and Exploratory Prototyping = 157
• 3.2. ES Development is Large Scale = 159
• 4. SE Solutions to Accidental Problems in ES Development = 160
• 4.1. Logical Modularity of the Process = 160
• 4.2. Functional Modularity of the Process = 161
• 4.3. Product Modularity = 163
• 4.4. Logical Modularity of Validation and Verification = 163
• 5. A New Paradigm for Expert System Development = 166
• 6. Example = 171
• 6.1. Problem Specification = 171
• 6.2. Solution Specification = 172
• 6.3. Design Specification = 174
• 6.4. Implementation = 174
• 6.5. Verification and validation = 176
• 7. Conclusions = 178
• Reference = 179
• Appendix = 186
• Chapter 7 : KNOWBEL : NEW TOOLS FOR EXPERT SYSTEM DEVELOPMENT / H. Wang, J. Mylopoulos, A. Kushniruk, B. Kramer and M. Stanley
• 1. Introduction = 189
• 2. Background = 189
• 2.1. Classification of Expert System Building Tools = 190
• 2.2. Knowledge Representation in Commercial Expert System Tools = 192
• 2.3. Experimental Knowledge Representation Systems = 195
• 3. Development of KNOWBEL = 197
• 3.1. System Architecture = 197
• 3.2. The Telos Level = 198
• 3.3. The MRS Level = 204
• 3.4. The Temporal Reasoner = 207
• 3.5. Reimplementation of KNOWBEl = 210
• 3.6. AP Hypertext User Interface = 212
• 4. Aplications = 218
• 4.1. Expert Systems for Prcess Control : The APACS Project = 218
• 4.2. Intelligent User Interfaces = 223
• 5. Conclusions = 225
• References = 227
• Chapter 8 : COOP : A SELF-ASSESSMENT BASED APPROACH TO COOPERATING EXPERT SYSTEMS / S. Shekhar and C.V. Ramamoorthy
• 1. Introduction = 231
• 2. Model of Cooperation = 233
• 3. Environment to Support C.E.S. Applications = 234
• 4. Annotated PROLOG = 236
• 4.1. Semantics of Annotaetd Prolog Programs = 238
• 4.2. Semantics of APL = 239
• 4.3. Proof Procedure for APL = 240
• 4.4. Implementation = 241
• 4.5. Sensitivity Analysis = 242
• 5. Yellow Pages : Locating Relevant Expertise = 243
• 5.1. Organizing the Knowledge = 244
• 5.2. Implementation = 245
• 6. How to Cooperate : Policies and Mechanisms = 248
• 6.1. Cooperation PoliCies= 249
• 6.2. Cooperation Mechanisms = 249
• 7. Network Event Manager = 251
• 8. Worm Mechanism = 252
• 8.1. Design = 253
• 8.2. Implementation = 254
• 9. Status and Future Work = 255
• References = 256
• Chapter 9 : EXPERIENCES MADE USING THE EXPERT SYSTEM SHELL G2 / S. Nilsen
• 1. Introduction = 259
• 2. The General Need for Expert System Technology = 260
• 3. Graphical Data Structures = 262
• 4. The SAS-Ⅱ System = 262
• 5. Requirements to the SAS-ⅡSystem = 264
• 6. Basic G2 = 267
• 7. Experiences Applying G2 for the SAS-ⅡProject = 268
• 7.1. The Applicability of the G2 Tool = 268
• 7.2. Structuring and Accessing the Knowledge Base = 271
• 7.3. Getting Data Into and Out from SAS-Ⅱ = 273
• 7.4. Time Efficiency and Stability = 273
• 7.5. Adaptability to Future Work = 274
• 8. Conclusion = 274
• References = 275
• SECTION 3 : KNOWLEDGE AND INFORMATION BASED SYSTEMS
• Chapter 10 : KEDE - A HYBRID KNOWLEDGE ENGINEERING DEVELOPMENT ENVIRONMENT / Z. Zheng and W. Li
• 1. Introduction = 279
• 1.1. Spin-off Tools = 282
• 1.2. Purpose Tools = 283
• 1.3. Knowledge Engineering Languages = 284
• 1.4. Hybrid Toolkits = 286
• 2. Knowledge Representation = 289
• 2.1. Frames = 289
• 2.2. semantic Nets = 300
• 2.3. Logic = 301
• 2.4. Procedural Knowledge = 304
• 2.5. Object Oriented Knowledge Representation = 304
• 3. Relations = 305
• 3.1. System-Defined Relations and Inheritance = 305
• 3.2. User-Defined Relations = 307
• 4. Data Oriented Programming = 308
• 5. Object-Oriented Programming = 311
• 5.1. The Classes and Instances = 311
• 5.2. The Methods = 311
• 5.3. Message Sending = 312
• 6. Logic-Oriented Programing = 314
• 6.1. Backward Chaining = 314
• 6.2. Forward Chaining = 315
• 6.3. Combining Forward and Backward Chaining = 316
• Conclusions = 316
• References = 317
• Chapter 11 : KNOWLEDGE AS A KEY COMPONENT IN INTELLIGENT INFORMATION SYSTEMS / M. P. Papazoglou
• 1. Introduction = 321
• 2. Active Information Systems = 323
• 2.1. The AIS Substrates = 323
• 2.2. Knowledge Representation = 324
• 2.3. Taxonomy of Knowledge = 327
• 3. Organizational Knowledge = 327
• 3.1 Intensional and Extensional Knowledge = 328
• 4. Organizational Knowledge Querying = 329
• 4.1. Querying the Factual Level = 329
• 4.2. Querying the Conceptual Level = 330
• 5. Concluding Remarks and Further Research = 338
• References = 340
• Chapter 12 : WIDE SCALE DEPLOYMENT OF KNOWLEDGE BASED SYSTEMS / Q. O'Neal
• 1. Introduction = 344
• 2. Examples of Leverage from Wide Scale Deployment = 345
• 3. The First Prerequisite to Propagation : Original Success = 345
• 3.1. KBS Project Planning = 345
• 3.2. KBS Application Implementation = 350
• 4. The KBS Project Database = 351
• 5. Establishing a Climate of Acceptance = 351
• 5.1. Preventing the "Not Invented Here" Syndrome = 352
• 5.2. KBS Language or Shell Selection = 352
• 5.3. Natural(Human) Language = 353
• 5.4. Project Awareness Vehicles = 353
• 6. Distribution and Introduction = 353
• 6.1 Propagation Types = 353
• 6.2 Distribution = 354
• 6.3. Introduction at the Receiving location = 354
• 7. Maintaining and Extending the Propagated application = 354
• 8. The Role of an Enterprise-wide Support Organization = 355
• 9. The Propagation Checklist = 355
• 10. Conclusion = 355
• References = 357
• Appendix A. KBS Propagation Checklist = 358
• Appendix B. Mini Business Case Project Qualification Process Form = 360
• Chapter 13 : A CANONICAL MODELING FACILITY FOR CAPTURING KNOWLEDGE IN INFORMATION BASE INTEGRATION / L. Marinos and J.Lee
• 1. Introduction = 361
• 2. Issues = 363
• 3. Existing Approaches = 364
• 4. The Proposed Approach = 367
• 4.1. Centralized Clusters Architecture = 367
• 4.2. The Modeling Facility of the Global System = 371
• 4.3. A Global Entity Example = 374
• 4.4. The Functions of the Global Manager = 377
• 5. Conclusion = 380
• References = 380
• Chapter 14 : HIGH LEVEL REASONING IN COOPERATIVE KNOWLEDGE SYSTEMS / F. Zheng and D. Liu
• 1. Introduction = 383
• 2. The Bsic Description of Multi-Dimentsio Argument = 386
• 3. The Calculation of Dependence Value = 390
• 4. The Calculation of External Constraints = 394
• 5. The Introduction of Multi-Dimentsion Argument System Framework PAT-1 = 395
• 6. Conclusion = 397
• References = 397
• Chapter 15 : A DATA PARALLEL SHELL FOR LARGE KNOWLEDGE BASES / A. K. Bansal and J. L. Potter
• 1. Introduction = 399
• 1.1. Exploiting Data Parallelism - An Associative Solution = 402
• 2. Background - Conventional Implementation = 405
• 2.1. Data Structures in Logic Program = 407
• 3. Problems with Conventional Models = 407
• 3.1. Problems with Data Representation = 407
• 3.2. Problems with Control Flow = 408
• 3.3. What are the Alternatives = 409
• 4. Associative Computing and data Parallelism = 410
• 4.1. ASC - An Associative Computing Language= 410
• 5. Data Parallel Prolog = 414
• 5.1. Associative Logical data structures = 414
• 5.2. Representing Logic programs Associatively = 415
• 6. Goal Reduction = 416
• 6.1. Pruning the Undesired Clauses = 418
• 6.2. Identifying Potential Bindings using Data Parallelis = 419
• 6.3. Determining Matched Data Clauses = 421
• 6.4. Associative Unification and Its Usage = 421
• 6.5. Handling Bindings = 422
• 7. Data Parallel Prolog Shell = 423
• 7.1. Model Structure = 424
• 7.2. Environment Management = 424
• 8. Model Behavior = 426
• 8.1. Handling Goal Reduction = 426
• 8.2. Handling Forward Control Flow = 427
• 8.3. Handling Backtracking = 428
• 8.4. Handling Cuts = 428
• 8.5. A Simple Example = 428
• 8.6. System Predicates and Extra Logical Predicates = 429
• 9. Conclusion = 429
• References = 430
• Chapter 16 : A KNOWLEDGE BASED APPROACH FOR THE SPECIFICATION AND ANALYSIS OF REAL-TIME SOFTWARE SYSTEMS / J.J.P. Tsai and H.-C. Jang
• 1. Introduction = 432
• 2. Real Time Requirements Specification Language = 434
• 2.1. Real-Time System = 434
• 2.2. Real-Time process = 434
• 2.3. Timing Constraint = 435
• 3. RT-FRORL System Model = 436
• 3.1. Modeling Periodic Process Using RT-FRORL = 437
• 3.2. Modeling Sporadic Process Using RT-FRORL = 437
• 3.3. Example = 438
• 4. Formal Foundation of RT-FRORL = 439
• 5. Converting RT-FRORL Specification to Its Underlying Logic Formula = 443
• 6. Requirements Analysis in RT-FROLRL = 444
• 7. Conclusion and Future Research = 448
• References = 449
• SECTION 4 : EVALUATION OF KNOWLEDGE BASES
• Chapter 17 : A TOOL FOR KNOWLEDGE BASE VERIFICATION / D. Zhang and D. Nguyen
• 1. Introduction = 455
• 2. Related Work = 456
• 3. A Verification Tool = 458
• 3.1. Inconsistencies and Incompleteness = 458
• 3.2. Pr/T Net Model of Rule Base = 461
• 3.3. Syntactic Pattern Recognition = 468
• 4. Implementation = 469
• 4.1. Components of PREPARE = 469
• 4.2. Transformer = 472
• 4.3. Scanner = 472
• 4.4. Formulator = 475
• 4.5. Classifier = 476
• 5. Discussions = 483
• 6. Conclusion = 485
• References = 485
• Chapter 18 : PRODUCTION GRAPH : A GRAPH THEORETICAL MODEL FOR CHECKING KNOWLEDGE BASE ANOMALIES / E.L. Lim, J. McCallum and K.H. Chan
• 1. Introduction = 487
• 1.1. Anomalies in Rule Based Systems = 489
• 1.2. Knowledge Debugging Tools = 490
• 1.3. Considerations for Anomaly Checking = 494
• 1.4. Redefinitions of the Anomalies = 495
• 1.5. Production-Graph = 495
• 1.6. A Guide to this Article = 496
• 2. A Restricted First Order Language for P-Graph = 496
• 2.1. A Restricted First Order Lnaguage = 497
• 3. Basic Elements of P-Graph = 500
• 3.1. A Toy Design Expert System = 500
• 3.2. Considerations in Developing P-Graph = 502
• 3.3. Formal definition of P-Graph = 505
• 3.4. Extension of P-Graph = 509
• 4. Structures of P-Graph for Anomaly Checks = 511
• 4.1. P-Graph Structures = 511
• 4.2. Inference Mechanism of P-Graph = 513
• 4.3. Advantages of P-Graph = 513
• 5. Anomalies : Definition and Detectio using P-graph = 514
• 5.1. Conflicting Facts and Rules = 514
• 5.2. Incompleteness - Missing Rules = 515
• 5.3. Redundant Rules = 516
• 5.4. Subsumed Rules = 517
• 5.5. Unreachable Rules = 518
• 5.6. Deadend Rules = 519
• 5.7. Cyclic Rules = 519
• 6. Discussion = 520
• References = 521
• Chapter 19 : VALIDATION OF NONMONOTONIC KNOWLEDGE-BASED SYSTEMS / C. L. Chang, R. A. Stachowitz and J. B. Combs
• 1. Introduction = 524
• 2. Nonmonotonic Knowledge-Based Systems = 526
• 3. Validation of Nonmonotonic Knowledge-Based Systems = 531
• 3.1. Redundancy Checking = 531
• 3.2. Inconsistency Checking = 533
• 3.3. Incompleteness Checking = 534
• 4. Concluding Remarks = 535
• References = 536