Transcript
Knowledge Reuse and Agile Processes: Catalysts for Innovation Amit Mitra TCS, Global Consulting Practice, USA Amar Gupta University of Arizona, USA
INFORMATION SCIENCE REFERENCE Hershey • New York
Acquisitions Editor: Development Editor: Senior Managing Editor: Managing Editor: Copy Editor: Typesetter: Cover Design: Printed at:
Kristin Klinger Kristin Roth Jennifer Neidig Sara Reed April Schmidt Cindy Consonery Lisa Tosheff Yurchak Printing Inc.
Published in the United States of America by Information Science Reference (an imprint of IGI Global) 701 E. Chocolate Avenue, Suite 200 Hershey PA 17033 Tel: 717-533-8845 Fax: 717-533-8661 E-mail:
[email protected] Web site: http://www.igi-global.com and in the United Kingdom by Information Science Reference (an imprint of IGI Global) 3 Henrietta Street Covent Garden London WC2E 8LU Tel: 44 20 7240 0856 Fax: 44 20 7379 0609 Web site: http://www.eurospanonline.com Copyright © 2008 by IGI Global. All rights reserved. No part of this publication may be reproduced, stored or distributed in any form or by any means, electronic or mechanical, including photocopying, without written permission from the publisher. Product or company names used in this set are for identification purposes only. Inclusion of the names of the products or companies does not indicate a claim of ownership by IGI Global of the trademark or registered trademark. Mitra, Amit, 1949Knowledge reuse and agile processes : catalysts for innovation / Amit Mitra & Amar Gupta. p. cm. Summary: "This book addresses flexibility of both business and information systems through component technology at the nexus of three seemingly unrelated disciplines: service-oriented architecture, knowledge management, and business process management. It provides practitioners and academics with timely, compelling research on agile, adaptive processes and information systems, and will enhance the collection of every reference library"--Provided by publisher. Includes bibliographical references and index. ISBN-13: 978-1-59904-921-2 (hardcover) ISBN-13: 978-1-59904-023-3 (ebook) 1. Knowledge management. 2. Management information systems. 3. Business logistics--Data processing--Management. 4. Computer network architectures. 5. Information theory. I. Gupta, Amar. II. Title. HD30.2.M58 2008 658.4'038--dc22 2007024489 British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library. All work contributed to this book set is original material. The views expressed in this book are those of the authors, but not necessarily of the publisher.
If a library purchased a print copy of this publication, please go to http://www.igi-global.com/reference/assets/IGR-eAccess-agreement. pdf for information on activating the library's complimentary electronic access to this publication.
Dedication
I dedicate this book to my late mother, Sevati Mitra, for her unflagging encouragement. She would have been a proud and delighted mother had she lived to see the book in print. I also dedicate this book to my wife Snigdha and my children, Tanya and Trishna, for their understanding and support as I spent long hours of time that were rightfully theirs to create this book. Amit Mitra
I dedicate this book to my parents-in-law, Mr. Ram Roop Gupta and Mrs. Pushpa Guta, on their 50th wedding anniversary. They have been a tremendous source of encouragement and inspiration for me. I owe a lot to them! Amar Gupta
Table of Contents
Foreword............................................................................................................................................... ix Preface.................................................................................................................................................. xii Acknowledgment................................................................................................................................ xiv Chapter I. Introduction to This Book ................................................................................................. 1 Agility and the Problem of Change .................................................................................................. 1 Scope of This Book ........................................................................................................................... 4 The 24 Hour Knowledge Factory and the Semantic Web................................................................. 5 Service Oriented Architecture .......................................................................................................... 6 Other Approaches............................................................................................................................. 6 Supplementary Materials and Organization of This Book ............................................................... 7 Chapter II. Introduction to Structure of Knowledge ........................................................................ 9 Introduction To Knowledge ............................................................................................................ 10 Modeling the Real World................................................................................................................ 12 Metaworld of Information .............................................................................................................. 12 The Repository of Meaning ............................................................................................................ 24 The Problem of Perspective............................................................................................................ 28 Chapter III. The Architecture of Knowledge ................................................................................... 34 The End of Common Sense: Hidden Chaos in the Heart of Complexity........................................ 34 The Architecture of Knowledge ...................................................................................................... 43 Chapter IV. The Pattern At the Root of It All .................................................................................. 67 Measure of Similarity: The Proximity Metric................................................................................. 68 The Ontology of Information Space ............................................................................................... 69 Properties of Patterns In Information Space.................................................................................. 76 Domains of Meanings vs. Format .................................................................................................. 81 The Object and the State Machine ............................................................................................... 108 Chapter V. Relationships .................................................................................................................. 121 Inverse of A Relationship.............................................................................................................. 122 Recursion and Reflexivity ............................................................................................................. 130 Idempotency.................................................................................................................................. 131
Symmetrical, Asymmetrical, and Antisymmetrical Relationships ................................................ 131 The Order and Degree of Relationships....................................................................................... 132 The Cardinality Ratio of Relationships ........................................................................................ 134 The Cardinality Ratios of Bijective and Surjective Relationships................................................ 134 Cardinality and Other Properties of Higher Order, Higher Degree Relationships ..................... 135 Mutual Inclusion and Exclusion of Relationships........................................................................ 140 The Cardinality of Subtypes ......................................................................................................... 141 Instance Level Constraints On Cardinality.................................................................................. 143 Compositions of Relationships ..................................................................................................... 143 The Capacity For Relationships................................................................................................... 148 Transitivity, Atransitivity, and Intransitivity................................................................................. 149 Collections of Objects and the State Space of Relationships ....................................................... 152 Slicing and Dicing Associations Between Objects ....................................................................... 152 Chapter VI. Object Aggregation...................................................................................................... 164 Emergent Properties of Aggregate Objects .................................................................................. 166 The Information Content of Aggregate Objects ........................................................................... 167 The Information in Aggregation vs. the Information in Composition .......................................... 171 Location, Containment, and Incorporation.................................................................................. 173 Chapter VII. Processes, Events, and Temporal Relationships...................................................... 178 Resources and Work Products ...................................................................................................... 179 Cycle Time .................................................................................................................................... 181 Temporal Inverses, Reversibility, and Reversion ......................................................................... 181 Temporal Recursion, Temporal Reflexivity, and Temporal Idempotency ..................................... 182 Temporal Asymmetry .................................................................................................................... 183 Temporal Mutability ..................................................................................................................... 184 Temporal Order ............................................................................................................................ 185 Temporal Degree .......................................................................................................................... 185 Temporal Cardinality: Concurrency, Repeatability, and Batch Processes .................................. 187 Efficiency and Productivity........................................................................................................... 189 Capacity For Temporal Relationships.......................................................................................... 189 Governance and Nonstationarity ................................................................................................. 190 Events ........................................................................................................................................... 191 Succession Constraints: Temporal Relationships Between Events .............................................. 192 The Metamodel of Relationship.................................................................................................... 255 Chapter VIII. Crossing the Chasm: Business Process To Information Systems......................... 289 Transforming Business Processes into Effects of Events ............................................................. 290 Transforming Business Processes into Information Systems Control Processes ......................... 292 Transforms that Implement Non-Temporal Relationships............................................................ 297 The Operation of Effects............................................................................................................... 300 Information Input-Output Processes ............................................................................................ 304 When Rules are Violated............................................................................................................... 306
Chapter IX. The Nature of Constraints .......................................................................................... 314 The Shaping of Objects................................................................................................................. 314 Patterns of Perspective and the Metamodel of Constraint........................................................... 316 Chapter X The Whole Shebang: The Integrated Metamodel of Knowledge .............................. 322 What is the Model of Knowledge and Why is it Useful? .............................................................. 322 Methodology................................................................................................................................. 325 The Integrated Model of Knowledge ............................................................................................ 326 Appendix I. Semantics of Pattern.................................................................................................... 333 Appendix II. Notes ............................................................................................................................ 338 Appendix III. Suggested Reading.................................................................................................... 366 Appendix IV. Meanings, the Semantic Web, Ontology, OWL and RDF...................................... 392 About the Authors............................................................................................................................. 406 Index................................................................................................................................................... 407
Detailed Table of Contents
Foreword............................................................................................................................................... ix Preface.................................................................................................................................................. xii Acknowledgment................................................................................................................................ xiv Chapter I. Introduction to This Book ................................................................................................. 1 Chapter II. Introduction to Structure of Knowledge ........................................................................ 9 This chapter defines knowledge and the need to coordinate knowledge, discusses why this is difficult and how the concept of normalization of knowledge can help coordination and agility; introduces the concept that knowledge has a structure and how it consists of indivisible components called “atomic rules”; describes how business processes and services are derived from atomic rules; introduces the modeling of behavior and the multiple perspectives related to the assembly of knowledge Chapter III. The Architecture of Knowledge ................................................................................... 34 Introduces the layered structure of knowledge and describes why chaos rides on the wings of change; illustrates why traditional approaches risk chaos and unintended side effects when the complexity and scope of business processes or information systems exceed a critical threshold Chapter IV. The Pattern At the Root of It All .................................................................................. 67 Defines how patterns are the basis of knowledge and measurability; introduces the concept of “information space,” in which patterns of pure information create meanings; distinguishes meanings from their physical representations and establishes the equivalence of objects and patterns; demonstrates how joining and constraining meanings creates new patterns of information, which are new meanings and hence the ability to configure meanings from other meanings; shows how this provides the basis for assembling knowledge from components
Chapter V. Relationships .................................................................................................................. 121 Shows how interactions between objects create new meanings; develops a model for business rules and shows how mutability supports innovation; introduces the rules that manipulate patterns of information to support inference and innovation Chapter VI. Object Aggregation...................................................................................................... 164 Describes meanings that emerge from aggregation; shows how concepts like containment and subtyping are configured from the concept of location Chapter VII. Processes, Events, and Temporal Relationships...................................................... 178 Describes business processes; shows how structured and unstructured business processes morph out of relationships; integrates business rules, processes, and inference into a single holistic information based ontology of meaning, tying them to business goals and agility; describes how product and process innovation may be partially automated and how processes and goals may be engineered to support business agility Chapter VIII. Crossing the Chasm: Business Process To Information Systems......................... 289 Describes the information architecture that transitions business semantic into computational processes; provides the information architecture of the interface between business semantic and automation Chapter IX. The Nature of Constraints .......................................................................................... 314 Describes how inchoate information is carved into normalized meanings and properties by constraints; describes the essential identity between a law and its outcome Chapter X The Whole Shebang: The Integrated Metamodel of Knowledge .............................. 322 Describes how the information in this book relates to that in its companion books and summarizes the conclusions that flow from the integrated model of knowledge. This chapter also shows, with several examples, how the entire scheme is integrated into one unified context and overarching high level structure of information, which leads to the concept of knowledge itself Appendix I. Semantics of Pattern.................................................................................................... 333 Appendix II. Notes ............................................................................................................................ 337 Appendix III. Suggested Reading.................................................................................................... 366 Appendix IV. Meanings, the Semantic Web, Ontology, OWL and RDF...................................... 392 About the Authors............................................................................................................................. 406 Index................................................................................................................................................... 407
ix
Foreword
New technologies, like new ideas, take time to become established. When they are first presented, they are met with a mix of enthusiasm and skepticism. Once tried, if success is not immediate—and it hardly ever is—those who opposed the innovation are quick to point out that they said the innovation would never work. Later, after the idea and the culture have had time to get to know one another and the new idea or technology is understood better, it often begins to flourish. The idea of describing business processes as knowledge networks and sets of rules began in the 1980s with what were then called expert or knowledge systems. The first expert systems used rules to capture the knowledge of business experts and then made that knowledge available to other experts by putting the rules into a software system that, given information about a specific problem, could make an expert-level recommendation. As the early expert systems got larger, it was determined that rules alone were too clumsy. Hence, by the mid-1980s, most of the more sophisticated expert system-building tools incorporated objects (they were called Frames in those days). In essence, the objects in a sophisticated expert system-building tool formed a network that described the vocabulary of problem, and rules were added to reason about the facts as they were accumulated by the system. When one used these more sophisticated expert system-building tools, one began by accumulating knowledge from experts. Thus, if one wanted to build an expert system to assist with home loans, one would begin by working out the vocabulary of loans. One would probably identify vocabulary objects like Home, Payment, Credit, Interest, Calendar, and so forth. Payments would probably have attributes like down payment and monthly payment, while Credit might have attributes like income, credit history, and so forth. In other words, one would construct a cognitive model of all of the concepts or words that a loan officer typically used—questions, in effect, that the loan officer would ask. Then, one would begin to add rules that could reason with the information one had about a specific case. For example: If the individual’s credit history was superior, and her salary was $130,000 a year, and she could make a down payment of $50,000, what type of loan would she qualify for? The objects and rules formed an abstract model of the concepts and rules an expert could use to organize knowledge about a particular subject and to reason about it to reach conclusions. By the end of the 1980s, most companies had given up on expert systems. They concluded that expert systems could be built but that the knowledge in the systems degraded too rapidly. One could capture human expertise in an expert system, but the system quickly became obsolete. Real human experts are constantly learning, reading journals, talking with colleagues about their latest experience, and attending conferences. They are constantly updating the knowledge structures and rules they use to analyze and solve problems. Thus, the problem with expert systems was not in the construction but in the maintenance.
x
It was easier to keep the expert, because the system that would replace him required that you keep him anyway, to maintain the expert system. This might have been the end of the idea that rules could be useful, but, in fact, it was only the beginning. Individuals who had learned about rules while building expert systems quickly realized that they could build systems to capture and automate more mundane human decisions—those based on welldefined corporate policies. Policies and associated business rules were easier to capture and changed less frequently. Thus, the interest in expert systems in the 1980s mutated into an interest in Business Rules in the 1990s and that application of rule technology is now flourishing. Many financial companies have large business rule groups that are responsible for defining and managing the business rules used throughout their organizations. At the same time that the business rules movement was showing how business rules could be used in practical situations, others were exploring patterns, business processes, and automated software tools that support business process modeling. Today, business rules and business processes are being integrated in new and creative ways. Amit Mitra and Amar Gupta propose to apply what I think of as a mixture of the expert systems approach to business process modeling and to the now popular business rules approach. In essence, they would build object models that described the vocabulary and business rules of an area of business—say Financial Management. If one then sought to create a business process in the area of Financial Management, one would, in essence, create process objects that would inherit information from the more generic Financial Management model. Mitra and Gupta refer to their high-level constructs as reusable patterns of business knowledge. They have written three books explaining this approach. This is the third. In the first, Agile Systems, they proposed a Universal Pattern that includes objects like Event, Fund, Energy, Physical Object, Person or Organization, Place, and Information. They work out the basic attributes of these objects and define some of the rules or constraints that apply to them. Then they start to create submodels for more specialized business activities. They consider, for example, a shipment and transportation cluster, a document and information cluster, a task-resource cluster, a meeting and agreement cluster, and a buying and selling cluster. Mitra and Gupta went on to propose that companies consider creating a knowledge machine. In essence, it would be a huge expert system that had all the knowledge of all the terms used by businesses and all the critical constraints or business rules. Anyone with a specific process problem would define the process, determine what elements of the process inherited what vocabulary, and instantly get an analysis of all the considerations and rules that might apply. To provide a foundation for the knowledge machine, Mitra and Gupta have explored all the technical problems one faces in creating this type of inheritance hierarchy. This kind of system cannot rely on the simple inheritance one finds in simpler object-oriented languages. It requires that one object can inherit from multiple parents and that some objects can inherit some features but not others from a given parent. These are programming problems that bedeviled the expert systems designers in the mid-1980s, and they still create technical and conceptual problems today. I mention this only to suggest that the first book is not light reading. It not only offers a survey of the high-level vocabulary and concepts of business but a survey of some very complex programming concepts as well. The second book, Creating Agile Business Systems with Reusable Knowledge, discusses the underlying ideas that form the foundation of the earlier book. This book probes the truly fundamental concepts involved, including the nature of reality and business, the nature of objects and attributes, and the meaning of domains. This book, the third in the series, describes how the underlying concepts described in the first book can be transformed into the business patterns described in the first book. Admittedly, the books were not
xi
published in what would seem to be the logical sequence, but now that all three are available, they can be read in whatever order the reader prefers. I found it easier to begin with the first book, which shows how everything fits together to create a business system and then to work back into the underlying theory once I understood why I would need it. Most, I suspect, will want to do that. Others may prefer to start with the first and then go to this volume that provides more on knowledge patterns and the automation of the business system. No matter where you begin, the journey will be challenging. It will also be rewarding if you really want to understand the potential for systematic, rule-based business systems analysis. These are ideas whose time is about to come, and this book and the other two in the series will give you the technical foundation and the vision to be ready when that time comes.
Paul Harmon is executive editor and founder of BPTrends. Harmon is a noted consultant, author, and analyst concerned with applying new technologies to real-world business problems. He is the author of Business Process Change: A Manager’s Guide to Improving, Redesigning, and Automating Processes (2003). He has previously co-authored Developing E-business Systems and Architectures (2001), Understanding UML (1998), and Intelligent Software Systems Development (1993).
xii
Preface
This book is part of a series of three complementary books (Figure P.1). The series addresses the pivotal issue of providing automated support for attaining business process resilience and information systems agility with little or no recurring manual intervention. The first two books, Agile Systems with Reusable Patterns of Business Knowledge: A Component Based Approach and Creating Agile Business Systems with Reusable Knowledge, were published by Artech House Press, Norwood, MA in October 2005 and Cambridge University Press, Cambridge, UK, in January 2007, respectively. The series as a whole addresses the basic organization of knowledge and how an integrated knowledge repository can be created from its shared components. This book, which is the final book of the series, addresses business rules and processes. In terms of content, Creating Agile Business Systems with Reusable Knowledge developed the semantics of Pattern and the concepts of Measurability, Distinction, Rules, Value, and Constraint, which are the basis of all knowledge. This book summarizes that foundation in Chapter IV and then builds upon it in subsequent chapters to provide additional depth. It addresses to a greater extent the components from which business rules and business processes are assembled and demonstrates how these components can automate reasoning and even some kinds of innovation. Each book is self-contained and may be read independently of the others.
Figure P.1. Reusing business knowledge: The three books CAPE tools and Value Chains start here. Our books start here...
Business Space
How
Cambridge University Press book
Transformation of abstract knowledge into reusable business processes
This book
Business Shared Business & Domain real World Knowledge Specification
Artech House book
Generation
The structure of Knowledge and Meaning
System
Design
Reusable patterns of knowledge
Requirements
What
Solution Space
CASE tools and Information Systems delivery start here.
xiii
The patterns that will be described in the following chapters facilitate the design of resilient services, business processes, and information systems. These patterns will also facilitate development of tools that can automate the design of “self aware” business services and adaptive information systems. The Semantic Web is a vision of the future, in which the World Wide Web operates on the plane of meanings. It envisions a future in which automation processes and integrates a World Wide Web of information based on the meanings of individual items of data. The patterns of information in this series of books describe meanings. These patterns do not need the Web to exist. However, they can be the cornerstone of the Semantic Web. The purpose of the semantics of knowledge we develop in this book is to normalize business rules and knowledge in order to reduce chaotic interactions and unintended side effects under the pressure of continual and rapid changes in scope, objectives, perspectives, and functionality. This book focuses on the concepts and models that integrate ontology, measurability, business rules, and business processes. The intent of this book is to anchor this integration in a cogent, overarching, nonstochastic model of knowledge and to demonstrate how such a model will result in agile and adaptable processes and information systems. Human, perceptual, and organizational issues, governance, and change management were addressed in the first book (Mitra & Gupta, 2005). This latest book discusses the risks associated with information quality and discusses processes for managing risks associated with violation of constraints. The long-term success of business is increasingly dependent on its underlying resilience and agility. Most analysis, methodologies, and traditional business process engineering practices place emphasis on operational efficiency and net profits at the expense of innovation and agility. However, innovation, agility, and coordination of information in support of value, from customers’ perspectives, are paramount in the global knowledge economy. In such an environment, research and processes that transcend departmental, corporate, and even national boundaries drive global excellence; innovation is not only supreme but is also made routine. This series of three books is tailored to support such an environment. The series demonstrates how new learning may be absorbed by flexible processes and information systems, which can be aligned automatically in lock step. The series supports the stated intent of the Object Management Group (OMG) to drive towards semantic integration of business rules, ontology, processes, and services in support of service orientation and self-aware business processes. The OMG has published its SBVR standard for business rules and is close to completing the BPDM model for business processes, which it eventually intends to integrate with SBVR. The Metamodel of Knowledge, in this series of books, supports OMG’s strategy by integrating the semantics that define business rules, business processes, reasoning, and shared knowledge. Read on to see how this can be done and discover the inhuman patterns of machine reasoning that will surprise you at the nexus of knowledge, process, and information.
xiv
Acknowledgment
We would like to say a very special thank you to Kristin Roth of IGI Global for her patience and guidance through the publishing process. We also thank Dane Sorenson, David Branson Smith, Katie James, Jessica Duran, Ruchika Agarwal, Julie Ovadia, and Trishna Mitra for their hard work, suggestions, and assistance in proofing and preparing the manuscript. We acknowledge, with sincere thanks, the valuable feedback provided by the reviewers. One of the authors (Amar Gupta) would like to thank his colleagues at the University of Arizona and at the Massachusetts Institute of Technology (MIT) for their support and encouragement.
1
Chapter I
Introduction to This Book
AGILITY AND THE PROBLEM OF CHANGE Change is difficult, complex, and risky because it has unintended side effects. Effects of change ricochet through systems via interactions between its parts. The larger the number of components, the more convoluted the system and the greater the chance of unintended side effects of change: more interactions imply a greater risk of multiple, complex impacts of change. Each impact has many consequences, which in turn will have many more until there is a cascading avalanche of changes, interactions, and impacts, which are difficult to manage, foresee, and qualify. This is the problem of change. The problem of change has persisted through 50 years of automation. Its solution has resisted every technology devised by man. In the beginning, our systems were small, simple, and of limited scope. However, automation opened up new opportunities to improve and integrate processes by coordinating ever larger numbers of elements in previously unanticipated ways. This meant coordination of information across continually broadening horizons, which led to processes that were more dependent on automated systems; further, these processes and systems were more complex, had even more interactions, and were therefore even riskier to change. Paradoxically, it
also led to the information economy, which thrives on change and innovation. We have created new technologies and methodologies at a prodigious rate to solve the burgeoning problem of change as our systems have evolved, matured, and integrated over the last 50 years. However, a solution to the problem of change has eluded us because every new approach has been the catalyst for the next level of complexity, which has then required a better, more sophisticated approach to managing change and innovation (Figure 1.1). Change impacts diverse business processes and cascades through multiple layers of the legacy information systems in a rapidly growing avalanche such that the initiator of the change is faced with the Hobson’s choice—either to make the change with huge overheads of cost, time, and risk, or to abandon the potential innovation because of the associated cost, time, and risk factors. The Y2K problem was a classic example. It cost the world around $600 billion (Yuen & Mitchell, n.d.) and exhausted a considerable part of the world’s professional resources, just to convert a two-digit representation of the calendar year to four digits1, which enabled automation to deal with the new millennium. As systems and processes became more integrated and tightly coupled, it became imperative to isolate and manage the effects of change. The strategy was to encapsulate densely clustered
Copyright © 2008, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.
Introduction to This Book
Figure 1.1. The evolution of information technology
Components Machine Assembler Functional Data Services 3GL Objects Hardware Code Code Decomposition Models BPM
1950 1960
1970
1980
1990
2000
20xx?
•Meanings are abstract patterns of information •Can we “normalize” meanings, configuring each from others?
interactions into components, which were coupled loosely with other similar components to produce requisite behaviors and outcomes. These components became the parts of more integrated, more modular systems across larger scopes, which were more maintainable because the impact of change was better managed within modules. This approach required abstraction of information. Each step of the journey in Figure 1.1 not only made business more agile and scalable but also led to higher levels of abstraction. The levels before it did not disappear; rather they hid themselves behind more malleable constructs that became the primary interface between man and machine or machine and machine. This helped the system to become more agile. As business processes became more tightly coupled with automation, the lack of agility in information systems became a serious bottleneck to product and process innovation. Several frameworks have attempted to solve this problem. Most have failed, or at best, have had very limited arguable success: Structured Programming, Reusable Code Libraries, Relational Databases, Expert Systems, Object Technology, CASE2 tools, code generators, and CAPE3 tools, to name a few. They failed because they did not adequately address the ripple effects of change—how business rules and knowledge may be represented so that we
2
may change a rule once and send corresponding changes rippling across all the relevant business processes. To do this, we need ontology, a schema of interrelated meanings, which are derived from each other. Ontology is a study of the meanings of things. It was a philosophical concept that became concrete and computable and, in so doing, took computation into the plane of meanings (see Appendix IV). It is the next advancement in the evolution of automation (see Figure 1.1). Currently, business rules are replicated in dissimilar formats in multiple, intermingled ways in multiple information systems and business processes. They must all be coordinated when any rule is changed. It makes change and innovation complex, perilous, and problematic to implement. This has been the most critical problem related to change.4 Purely technical approaches have failed miserably. Despite some claims to the contrary, the problem was not resolved in the 1950s when computer professionals replaced the intertwined programming code of machine language with assembly languages, or in the 1960s when the next generation of these professionals replaced the cumbersome code of assembly languages with that of third generation languages like COBOL and FORTRAN. During the 1970s and 1980s, it was not solved either, when the expert systems, relational databases, and CASE tools
Introduction to This Book
were deployed. In addition, in the 1990s, object technology was considered to be a panacea until tangled object inheritance became so much of a problem that many advocated making multiple inheritances illegal in tools of the day. Finally, in very recent years, as one hurtled towards business process management (BPM) and service oriented architecture (SOA) with their plug-and-play business services, the problem had not been solved either. This has happened because new and better automation triggered more tangling of these business rules. Therefore, the authors asked entirely different questions when initiating the research leading to this book: •
•
•
What is the natural structure of information that is used to represent business knowledge and services in a fully normalized and reusable form across diverse global business environments? What information is needed to model the stimulus response behavior of business processes and host organizations? If so many approaches have failed, why would a new one work?
The framework described in this book addresses these three issues by untangling business rules with an ontology derived from the inherent structure of information. By untangling business rules, even in complex legacy models and systems, one gains the unique capability to represent specific elements of business knowledge once, and only once, in a knowledge repository. Using this repository, the specific elements of knowledge can be designed to naturally manifest themselves, in appropriate forms, to suit the idiosyncrasies of different business contexts. Changes made at appropriate places will ripple through and impact relevant places within the concerned business systems with minimal or no human intervention. Not surprisingly, business professionals have long perceived that business information gained
in one context may be used in another situation. However, in order to attain this objective in automation, one must specify the knowledge with greater precision in the appropriate framework. This book addresses the quest to define a fundamentally reusable structure of knowledge in a language that can be understood by most business professionals and also by machines. These patterns of knowledge flow from theory and are validated by practice across a spectrum of different industries. Indeed, they had their genesis, not in abstract theory, but in the practical need to build the semantics of agile business for large diversified corporations, such as AIG and Verizon, in programs which one of the authors (Mitra) spearheaded. This series of three books provides a connection between the world of systems engineering and the world of business process engineering. It is a generalized framework that applies with equal ease to diverse industry and business applications, ranging from transportation to defense and agriculture to medicine. Figure 1.2 describes the overall scheme of the framework. The scheme in Figure 1.2 can assist in identifying reusable business services and predicting principal requirements, predicated on common patterns of knowledge, even before users articulate them. This can dramatically reduce the time needed to develop and to market new products and services. Moreover, the strategy can be a crucial asset in our intensely competitive business world, which increasingly depends on putting new ideas on the table in ever-shortening periods of time. Industry consortiums such as the OMG and W3C are developing standards for business rules and business processes (like SBVR for business rules and BPDM for business processes from OMG and RDF for metadata from W3C). The W3C also has a recommendation called OWL (see Appendix IV) for a limited part of the ontology layer at the apex of the pyramid in Figure 1.2. However, these models are not integrated yet and are therefore limited in their ability to
3
Introduction to This Book
Figure 1.2. The business knowledge engineering framework
METAMODEL OF BUSINESS RULES
METAMODEL OF KNOWLEDGE
Measurement & Reasoning
METAMODEL OF UPPER ONTOLOGY
METAMODEL OF BUSINESS PROCESS
Agile Systems with Reusable Patterns of Business Knowledge from Artech House Publishers
CROSS-INDUSTRY BUSINESS KNOWLEDGE
INDUSTRY BUSINESS KNOWLEDGE
represent integrated knowledge and reasoning that flow through all aspects of business. This book shows how this can be done and presents a cogent integrated model of knowledge. Corporations such as IBM are developing the industry models at the base of the pyramid in Figure 1.2. Some examples are the IAA model for insurance, IBM’s health care models, Telecordia’s TMN architecture for telecommunications, the Supply Chain Council’s SCOR model for manufacturing, and others. In terms of the theme in Figure 1.2, these industry models are integrated neither horizontally nor vertically. This limits their ability to orchestrate and reuse knowledge across the diversity of business partners that form modern extended enterprises. These are the very enterprises that are enabled by the World Wide Web and the global knowledge economy and have the potential to make quantum leaps in the value they bring to end users of products and services. To bring true integration, agility, and coordination to the information enabled extended enterprises of the 21st century, the cross-industry layer of knowledge is critical. It allows a firm to innovate and reinvent its product markets, coordinates across business partners, and enables the business-on-demand concepts, enabled by the Web,
4
which corporations like IBM have envisioned for the future. Making business systems entirely maintenance free is the ultimate vision. Systems based on software will automatically adapt to chaos and change. These systems will be assisted by automated intelligent agents that will hopefully, someday, maintain software and adapt to change even as it occurs. They might even anticipate change, and perhaps thrive on it, like the businesses of the 21st century, which they will support.
SCOPE OF THIS BOOK This book and the other two books in the series focus on normalization, encapsulation, and reuse of business knowledge across a broad spectrum of industries and dissimilar business functions.5 This book identifies the information that describes normalized knowledge. It does not describe the sequence of tasks that are required to capture this information (how the information is captured may vary widely). Thus, it is not a cookbook of sequenced activities to build components of normalized knowledge; rather it provides the foundation for cookbooks of that kind and a basis
Introduction to This Book
for evaluating how complete existing cookbooks are in terms of the information they must collect to model business knowledge. Although business knowledge and technology are considered independent entities, the knowledge that is embedded within processes must be supported by an array of technologies, both manual and automated in nature, in order to derive full benefits. Frequently, large organizations and extended enterprises that are in close partnership in a supply chain have difficulty in coordinating their processes. This leads to waste, inefficiency, lack of coordination, and loss of agility. Different divisions and units of these enterprises have different and sometimes confusing business rules. On closer analysis, it becomes apparent that these apparently different rules, manifested in different procedures, implemented in different systems, which might run on different technology platforms, are merely different expressions and implementations of the same generic rules. The dissimilar implementations are driven by different local legacies, characterized by their own geographical, technological, whimsical, political, and environmental parameters. It is possible to extract the shared business knowledge and intent from these diverse implementations. This shared knowledge focuses on the intent and semantics of the business. It is platform and procedure independent. It provides the basis for a shared “federated” business model. The federated model can coordinate the shared semantics of the business, which includes processes, rules, and information in the “federation” of businesses. If the federation wishes to reuse this shared knowledge, it must store it in an electronic repository. Although the federated model itself is technology independent, in the repository, it will be an array of information expressed explicitly on physical media in physical formats. It is thus an electronic artifact. We have named these artifacts Business Knowledge Artifacts, often abbreviated to Knowledge Artifacts, in this book.
Traditional software and hardware components differ from these Knowledge Artifacts. These Knowledge Artifacts are the components from which business knowledge and its semantics may be configured. New learning leads to adaptation by changing configurations of Knowledge Artifacts. This book identifies these Knowledge Artifacts and shows their relationship to software components. It also shows how automating these configurations can automate reasoning and the creation of the right processes for a business. As such, these Knowledge Artifacts encapsulate business intelligence as meanings and reasoning that can be stored as reusable components within an electronic knowledge repository.
THE 24-HOUR KNOWLEDGE FACTORY AND THE SEMANTIC WEB The purpose of the 24-Hour Knowledge Factory is to drastically reduce the time needed to develop information systems, and to facilitate effective knowledge-based processes. It is like a relay race that envisions a globally distributed work environment, in which global teams work on projects around the clock. Each team member works a normal workday in his or her time zone, and at close of business passes the baton to another member in a different time zone, who then continues the same task6. One of the authors (Gupta) has done extensive research on the concept and has successfully tested the efficacy of this approach in large industrial and academic environments. Knowledge artifacts will facilitate the operation of such a factory because they are the components from which business knowledge and its semantics are configured, coordinated, and used to automate the creation of information systems and services. New learning and other changes lead to adaptation by changing configurations of knowledge artifacts, and thereby changing the
5
