OPIS
The monograph is an extensive exploration of explainable artificial intelligence (XAI), specifically applied to corporate planning within the process industry, focusing on chemicals and pharmaceuticals. It outlines the economic significance of AI, differentiates between symbolic and subsymbolic (black-box) AI, and explains the necessity of XAI to enhance trust, transparency, and decision-making quality in AI systems. The core objective of the monograph is to develop a reference architecture, named Re_fish, which integrates symbolic and subsymbolic approaches for practical application in corporate planning scenarios.
SPIS TREŚCI
1
INTRODUCTION AND THE SCOPE OF THE BOOK (17)
1.1. Motivation and Relevance / 24
1.2. The Economic Perspective of AI / 37
1.3. The Microeconomic Perspective of AI / 49
1.4. Research Goal and Research Questions / 55
1.5. Research Theory and Design / 62
2
PLANNING IN THE PROCESS INDUSTRY (77)
2.1. Introduction / 77
2.2. The Process Industry / 79
2.2.1. The Specifics of the Process Industry / 80
2.2.2. Key trends of the Process Industry / 87
2.2.3. Challenges of the Process Industry / 93
2.2.4. AI Support in the Process Industry / 109
2.3. Planning and Decision-Making in the Process Industry / 121
2.3.1. Scenario Planning in the Process Industry / 126
2.3.2. Integrated Business Planning in the Process Industry / 131
2.3.3. Decision Making and Explanations in Planning in the Process Industry / 142
2.3.4. Stakeholders in Corporate Planning in the Process Industry / 168
2.4. Information Systems to Support Planning and Decision-Making in the Process Industry / 174
2.5. Classical Decision Support Systems, Business Analytics, Data Science, and Reporting / 175
2.5.1. Classical Decision Support Systems / 178
2.5.2. Business Analytics, Predictive, and Prescriptive Analytics /179
2.5. Data Science / 181
2.6. Summary / 181
3
EXPLAINABLE ARTIFICIAL INTELLIGENCE IN CORPORATE PLANNING (185)
3.1. Introduction / 185
3.2. The Technical Perspective of Artificial Intelligence / 188
3.2.1. Machine Learning and Deep Neural Networks / 189
3.2.2. Knowledge-Based Systems / 196
3.2.3. Neuro-symbolic AI / 201
3.3. Explainable Artificial Intelligence / 203
3.3.1. Explainable or Interpretable Machine Learning / 207
3.3.2. Knowledge-Enabled Systems of Explainable AI / 224
3.3.3. Neuro-symbolic Systems of Explainable AI / 240
3.4. Ethical AI, Law , and Regulatory Requirements of Explainable AI / 241
3.5. Mapping the Stakeholders and Their Requirements / 245
3.6. Summary / 247
4
DESIGN OF A REFERENCE ARCHITECTURE FOR EXPLAINABLE AI (249)
4.1. Introduction / 249
4.2. Theoretical Basis of Reference Architectures / 251
4.3. Methodology to Develop Reference Architectures / 254
4.3.1. Methods to Develop a Reference Architecture / 255
4.3.2. Phase A: Architecture Vision / 271
4.3.2.1. Establish the Architecture Project / 271
4.3.2.2. Stakeholders, Concerns, and Business Requirements / 272
4.3.2.3. Confirm and Elaborate Business Goals, Drivers, and Constraints / 273
4.3.2.4. Define Scope / 273
4.3.2.5. Confirm and Elaborate Architecture/Business Principles / 273
4.3.2.6. Develop Architecture Vision / 274
4.3.2.7. Summary of Phase A / 274
4.3.3. Phase B: Business Architecture / 275
4.3.3.1. Select Reference Models, Viewpoints, and Tools / 275
4.3.3.2. Conduct Formal Stakeholder Review / 275
4.3.3.3. Finalise the Business Architecture and Update ADD / 276
4.3.3.4. Summary of Phase B / 276
4.3.4. Phase C: Information System Architecture / 277
4.3.4.1. Select Reference Models, Viewpoints, and Tools / 277
4.3.4.2. Summary of Phase C / 278
4.3.5. Phase D: Technology Architecture / 279
4.3.5.1. Select Reference Model, Viewpoints, and Tools / 280
4.3.5.2. Develop Target Technology Architecture Description / 281
4.3.5.3. Summary of Phase D / 281
4.3.6. Phases E to H: Implementation of a Concrete Reference Architecture / 282
4.3.7. Summary of the Methodology to Develop Reference Architectures / 284
4.4. Summary / 284
5
DEVELOPMENT OF A REFERENCE ARCHITECTURE FOR EXPLAINABLE AI IN CORPORATE PLANNING (287)
5.1. Introduction / 287
5.2. Development of the Re_fish Reference Architecture / 288
5.2.1. Preliminary, Purpose, and Scope / 290
5.2.2. Architectures of Knowledge-Enabled AI Systems / 302
5.2.3. Gathering and Synthesis of the Requirements / 320
5.2.4. Re_fish Business Architecture / 326
5.2.5. Re_fish Application Architecture / 328
5.2.6. Re_fish Technology Architecture / 343
5.2.7. Re_fish Overall Architecture / 343
5.2.8. Re_fish Lifecycle Management / 346
5.2.9. Re_fish Opportunities and Solutions / 350
5.3. Evaluation of the Re_fish Architecture – Design Science Evaluation and Expert Survey / 351
5.4. Adjustment of the Re_fish Reference Architecture / 372
5.5. Summary / 377
6
SUMMARY AND OUTLOOK (379)
6.1. Limitations and Future Work / 387
REFERENCES (389)
GLOSSARY (431)
APPENDIX (437)
LIST OF FIGURES (444)
LIST OF TABLES 446)
LIST OF GRAPHS (447)
