Data Science for Entrepreneurship: Principles and Methods for Data Engineering, Analytics, Entrepreneurship, and the Society

Description

The fast-paced technological development and the plethora of data create numerous opportunities waiting to be exploited by entrepreneurs. This book provides a detailed, yet practical, introduction to the fundamental principles of data science and how entrepreneurs and would-be entrepreneurs can take advantage of it. It walks the reader through sections on data engineering, and data analytics as well as sections on data entrepreneurship and data use in relation to society. The book also offers ways to close the research and practice gaps between data science and entrepreneurship. By having read this book, students of entrepreneurship courses will be better able to commercialize data-driven ideas that may be solutions to real-life problems. Chapters contain detailed examples and cases for a better understanding. Discussion points or questions at the end of each chapter help to deeply reflect on the learning material. Preface Acknowledgments About the Book Contents About the Editors Contributors Editors and Contributors 1: The Unlikely Wedlock Between Data Science and Entrepreneurship 1.1 Introduction 1.2 Defining Data Science and Entrepreneurship 1.3 Towards a Definition of Data Entrepreneurship 1.4 Processes of Data Science and Entrepreneurship 1.4.1 The Data Science Process 1.4.2 The Entrepreneurial Process 1.4.3 Comparing Data Science and Entrepreneurial Processes 1.5 The Data Entrepreneurship Framework References I: Data Engineering 2: Big Data Engineering 2.1 Introduction: The Big Data Engineering Realm 2.1.1 Data Engineering Challenges in Theory and Practice 2.2 (Big) Data Engineering to Leverage Analytics 2.2.1 Value-Driven Big Data Engineering 2.2.2 Key Fabric of Data Engineering 2.2.2.1 Intelligent Enterprise Application Architecture (iA)2 2.2.2.2 Data Pipelines 2.2.2.3 Data Lakes and Data Warehouses 2.2.3 MLOps: Data Engineering (Finally) Meets AI/Machine Learning Take-Home Messages References 3: Data Governance 3.1 Introduction 3.2 Motivational Case Studies 3.2.1 SODALITE Vehicle IoT 3.2.2 SODALITE Clinical Trials 3.3 Data Governance in a Nutshell 3.4 Data Governance Dimensions 3.4.1 Data Principles 3.4.2 Data Quality 3.4.3 Metadata 3.4.4 Data Access 3.4.5 Data Life Cycle 3.5 Data Governance Structure 3.5.1 Executive Sponsor 3.5.2 Data Governance Council 3.5.3 Data Custodian 3.5.4 Data Steward 3.5.5 Data User Groups 3.6 Contemporary Data Governance 3.6.1 Big Data Governance 3.6.2 IoT Data Governance 3.7 Case Studies with Data Governance 3.7.1 SODALITE Vehicle IoT Architecture 3.7.2 SODALITE Clinical Trial Architecture Take-Home Messages References 4: Big Data Architectures 4.1 Introduction 4.2 Background 4.2.1 Key Attributes of Big Data Systems 4.2.2 From Structured Data to Semi-structured Data 4.3 Lambda Architecture 4.4 Kappa Architecture 4.5 SEI-CMU Reference Architecture References 5: Data Engineering in Action 5.1 Introduction 5.2 The ANITA Project for the Fighting of Cybercrime 5.2.1 Data Collection 5.2.2 ANITA Architecture 5.2.3 Data Extraction 5.2.4 Data Management and Analysis 5.3 The PRoTECT Project for the Protection of Public Spaces 5.3.1 Objectives of PRoTECT 5.3.2 PRoTECT and the Data Fusion Approach 5.4 The Beehives Project for the Quality of Urban Biodiversity 5.4.1 Problem Description 5.4.2 Objectives 5.4.3 Data Gathering 5.4.4 Big Data Analytics for Biodiversity 5.4.5 Systemic Change 5.4.6 Bringing It All Together: The IoT Beehive Stratified Architecture Take-Home Messages References II: Data Analytics 6: Supervised Machine Learning in a Nutshell 6.1 Introduction 6.2 Supervised Learning: Classification 6.2.1 Motivating Example: Credit Card Fraud Detection 6.2.2 An Overview of Classifiers 6.2.3 Evaluating a Classification Model 6.2.4 Designing a Pipeline for Machine Learning Classification 6.2.4.1 Data Mining 6.2.4.2 Data Preprocessing 6.2.4.3 Data Classification 6.3 Supervised Learning: Regression 6.3.1 Simple Linear Regression 6.3.2 Regression Methods: An Overview 6.3.3 Evaluating a Regression Model 6.3.4 Designing a Pipeline for Machine Learning Regression 6.3.4.1 Data Mining 6.3.4.2 Data Preprocessing Take-Home Messages References Further Reading 7: An Intuitive Introduction to Deep Learning 7.1 Brief Historical Overview 7.2 Datasets, Instances, and Features 7.3 The Perceptron 7.3.1 The Decision Boundary 7.3.2 The Delta Learning Rule 7.3.3 Strengths and Limitations of the Perceptron 7.4 The Multilayer Perceptron 7.4.1 Combining Decision Boundaries 7.4.2 The Generalized Delta Learning Rule 7.5 Deep Neural Networks 7.5.1 Combinations of Combinations of … Decision Boundaries 7.5.2 The Generalized Delta Learning Rule in Deep Networks 7.5.3 From Two- to High-Dimensional Feature Vectors 7.6 Convolution: Shifting a Perceptron Over an Image 7.6.1 The Basic Convolution Operation 7.7 Convolutional Neural Networks 7.7.1 Convolutional Layers 7.7.2 Pooling Layers 7.7.3 Combinations of Combinations of … Features 7.7.4 Dense Layers 7.7.5 From AlexNet to Modern CNNs 7.8 Skin Cancer Diagnosis: A CNN Application 7.8.1 Introduction 7.8.2 Data Collection and Preparation 7.8.3 Baseline and Multitask CNN 7.8.4 Experiments and Results 7.8.5 Conclusion on the CNN Application References 8: Sequential Experimentation and Learning 8.1 Introduction 8.2 The Multi-Armed Bandit Problem 8.3 Solutions to Bandit Problems: Allocation Policies 8.3.1 ϵ-First 8.3.2 ϵ-Greedy 8.3.3 Upper Confidence Bound Methods 8.3.4 Thompson Sampling 8.3.5 Bootstrapped Thompson Sampling 8.3.6 Policies for the Contextual MAB Problem 8.4 Evaluating Contextual Bandit Policies: The Contextual Package 8.4.1 Formalization of the cMAB Problem for Its Use in Contextual 8.4.2 Class Diagram and Structure 8.4.3 Context-Free Versus Contextual Policies 8.4.4 Offline Policy Evaluation with Unbalanced Logging Data 8.5 Experimenting with Bandit Policies: StreamingBandit 8.5.1 Basic Example 8.5.2 StreamingBandit in Action References 9: Advanced Analytics on Complex Industrial Data 9.1 Introduction 9.2 Data Analytics for Fault Diagnosis 9.2.1 Maintenance of Equipment 9.2.2 Preparing the Data 9.2.3 Machine Learning Classifiers 9.2.4 Deep Learning Techniques 9.2.5 Fault Diagnosis in Practice 9.2.6 Simulating a Real-World Situation 9.2.7 Summary 9.3 Graph Signal Processing (GSP) 9.3.1 GSP Background 9.3.2 GSP Applications 9.3.3 Summary 9.4 Local Pattern Mining on Complex Graph Data 9.4.1 Overview 9.4.2 Local Pattern Mining on Graphs 9.4.3 Local Pattern Mining on Attributed Graphs 9.4.4 MinerLSD: Local Pattern Mining on Attributed Graphs 9.4.5 Application Example 9.4.6 Summary Take-Home Messages References 10: Data Analytics in Action 10.1 Introduction 10.2 BagsID: AI-Powered Software System to Reidentify Baggage 10.2.1 Business Proposition 10.2.2 System Overview 10.2.3 AI Engine 10.2.4 Software Engineering Aspects 10.3 Understanding Employee Communication with Longitudinal Social Network Analysis of Email Flows 10.3.1 Digital Innovation Communication Networks 10.3.2 The Relational Event Modeling Framework 10.4 Using Vehicle Sensor Data for Pay-How-You-Drive Insurance 10.4.1 Time Series 10.4.2 Driving Behavior Analysis References III: Data Entre 11: Data-Driven Decision-Making 11.1 Introduction 11.2 Introduction to Decision-Making 11.2.1 Decision-Making Characteristics 11.2.2 The Decision-Making Process and Decision Rules 11.2.3 Decision-Making for Entrepreneurs 11.3 Data-Driven Decision-Making 11.3.1 What Is Data-Driven Decision-Making? 11.3.2 Maturity Levels of Data-Driven Decision-Making 11.3.3 Methodology Options for Data-Driven Decision-Making 11.3.4 Data-Driven Decision-Making by Entrepreneurs 11.4 Data-Driven Decision-Making: Why? 11.4.1 Quality Reasons for Data-Driven Decision-Making 11.4.1.1 DDDM for Decision Quality 11.4.2 Capacity Reasons for Data-Driven Decision-Making 11.4.2.1 DDDM for Reducing Information Overload 11.4.3 Mental Reasons for Less Data-Driven Decision-Making 11.5 Data-Driven Decision-Making: How? 11.5.1 Overview of Data-Driven Decision-Making Solutions 11.5.2 Data-Driven Decision-Making Solutions for Programmed Decision-Making 11.5.2.1 Operations Research Solutions 11.5.2.2 Data Science Solutions 11.5.2.3 Recommender Systems 11.5.3 Data-Driven Decision-Making Solutions for Nonprogrammed Decision-Making 11.5.3.1 Agent-Based Modeling (ABM) 11.5.3.2 Case-Based Reasoning/Decision Analysis 11.5.3.3 Technology-Assisted Reviews (TAR) 11.5.3.4 Scenario-Based Decision-Making 11.5.3.5 Competitive Benchmarking References 12: Digital Entrepreneurship 12.1 Introduction 12.2 What Is Digital Entrepreneurship? 12.3 What Is Different in the Digital Economy? 12.3.1 How Do Digitization and Digital Artifacts Affect the Nature of Business and of New Venture Creation? 12.3.2 What Are the Implications for Entrepreneurship of the Nature of the Digital Economy? 12.4 Digital Platforms and Digital Entrepreneurship 12.4.1 Creating and Growing a Digital Platform Firm 12.4.2 Competing on Digital Platforms 12.5 Supporting and Regulating Digital Entrepreneurship 12.5.1 Understanding and Supporting Digital Entrepreneurial Ecosystems 12.5.2 Regulating Digital Entrepreneurship Take-Home Messages References 13: Strategy in the Era of Digital Disruption 13.1 Introduction 13.2 Disruption Driven by Business Model Innovations 13.2.1 Freemium Business Models 13.2.2 Sharing Economy Business Models 13.2.3 Usage-Based Business Models 13.3 Disruption Driven by Innovation Ecosystems 13.3.1 Supply-Side Synergies 13.3.2 Demand-Side Synergies 13.4 Disruption Driven by Platforms and Network Effects 13.5 Discussion 13.5.1 Trend 1: Industry Crossover Trends in a Digital World 13.5.2 Trend 2: Changing Competitive Landscape 13.5.3 Trend 3: Rising Customer Expectations References 14: Digital Servitization in Agriculture 14.1 Introduction 14.2 Servitization 14.3 Types of Services 14.4 Servitization in Agriculture 14.5 Digital Servitization 14.6 Digital Servitization in Agriculture References 15: Entrepreneurial Finance 15.1 Introduction 15.2 Pre-seed Financing and Support 15.2.1 Family, Friends, and Fools 15.2.2 Accelerators, Incubators, and Startup Studios 15.2.2.1 Incubators 15.2.2.2 Accelerators 15.2.2.3 Startup Studios: Venture Builders 15.3 Early Sources of Funding (Seed and Startup Stage) 15.3.1 Business Angels 15.3.2 Crowdfunding 15.3.3 Initial Coin Offerings (ICOs) 15.4 Venture Capital and Private Equity (Growth Stage) (Da Rin & Hellmann, 2019) 15.4.1 Ownership and Valuation 15.4.2 Preferred Shares 15.4.3 Staged Financing 15.4.4 Corporate Governance 15.4.5 Exit Routes 15.5 Tech Startup Financing in Practice 15.6 Answers to the Cases References 16: Entrepreneurial Marketing 16.1 Introduction 16.2 Defining Marketing and Sales 16.3 Customers Buy Solutions Rather Than Products 16.3.1 The Means-End Chain 16.3.2 Trade-Offs Regarding Radically New Products and Services 16.4 Co-developing and Positioning a New Product or Service 16.5 Organizing Customer Development as a Separate Process 16.6 A One-Page Marketing and Sales Plan 16.6.1 The General Motivation and Objectives 16.6.2 Three Main Pillars 16.6.3 Building a Marketing Information System 16.7 Leveraging Your Growing Customer Base 16.7.1 Segmentation and Targeting 16.7.2 Efficient A/B Testing of Value Proposition Take-Home Messages References IV: Data and Society 17: Data Protection Law and Responsible Data Science 17.1 Introduction 17.2 A Few Words on the Meaning of Privacy and Data Protection 17.3 Material Scope of Data Protection Law: Defining Processing and Personal Data 17.3.1 Defining Processing 17.3.2 Defining Personal Data 17.3.2.1 “Any Information” 17.3.2.2 “Relating to” 17.3.2.3 “Identified or Identifiable” 17.3.2.4 “Natural Person (Data Subject)” 17.3.3 Conclusion: Personal Data and Non-personal Data 17.4 Personal Scope of Data Protection: Controller and Processor 17.4.1 The Three Main Actors of Data Protection 17.4.2 Data Controllers 17.4.3 Data Processors 17.4.4 Problematic Situations 17.4.4.1 Controller or Processor? 17.4.4.2 Multiple Controllers 17.5 Art. 6, GDPR: The Need for a Legitimate Ground of Processing 17.5.1 Consent 17.5.1.1 Consent Must Be Given in Relation to a Specific Purpose 17.5.1.2 Consent Must Be Informed 17.5.1.3 Consent Must Be Unambiguous 17.5.1.4 Consent Must Be Free 17.5.1.5 Special Categories of Data: Explicit Consent 17.5.2 Contract 17.5.3 Vital Interests of the Data Subject 17.5.4 Performance of a Task Carried Out in the Public Interest or in the Exercise of Official Authority Vested in the Control 17.5.5 Compliance with a Legal Obligation to Which the Controller Is Subject 17.5.6 Legitimate Interests of the Data Controller or a Third Party 17.5.6.1 The Interest of the Data Controller: A Legitimate One 17.5.6.2 Interests or Fundamental Rights of Data Subject 17.5.6.3 Balancing of Interests Step 1: Qualify the Interests Step 2: Impact(s) on the Data Subject Step 3: Factors for Appraising the Impacts Step 4: Provisional Balance Step 5: Additional Safeguards Step 6: Final Balance 17.6 Art. 5 GDPR: Principles to Be Applied to the Processing of Data 17.6.1 Purpose Limitation Principle 17.6.1.1 Purpose Specification: Why? 17.6.1.2 Specific Purpose 17.6.1.3 Explicit Purpose 17.6.1.4 Legitimate Purpose 17.6.1.5 Different Purpose 17.6.2 Data Minimisation 17.6.3 Storage Limitation 17.6.4 Additional Obligations 17.6.4.1 Data Accuracy 17.6.4.2 Lawfulness, Fairness, and Transparency 17.6.4.3 Integrity and Confidentiality References 18: Perspectives from Intellectual Property Law 18.1 Introduction 18.2 Meeting the Criteria 18.2.1 The Formal Requirements of Copyright 18.2.2 Sui Generis Database Right 18.2.3 Trade Secret Right 18.2.4 Summary 18.3 The Scope of Protection 18.3.1 Copyright: Protected Subject Matter 18.3.2 Sui Generis Database Protection 18.3.3 Trade Secret Right 18.3.4 Summary 18.4 Exceptions and Limitations 18.4.1 Limitations of the Rights 18.4.2 Exceptions: Common Ground 18.4.3 Exceptions Specific to the Right 18.5 Alternative Sources Further Reading 19: Liability and Contract Issues Regarding Data 19.1 Introduction 19.2 General Characteristics of Private Law 19.3 What Is Data? 19.4 Contracts and Data 19.4.1 Formation of Contracts 19.4.2 Content of Contracts 19.4.3 Contractual Remedies 19.4.3.1 Prerequisites for Invoking a Remedy 19.4.3.2 The Available Remedies 19.5 Tort Law and Data 19.5.1 Fault Liability 19.5.2 Strict Liability 19.5.3 Causality and Defenses 19.5.4 Damages and Other Remedies in Tort Take-Home Messages References 20: Data Ethics and Data Science: An Uneasy Marriage? 20.1 Introduction 20.2 Data Ethics in Academia 20.2.1 Moral Theories 20.2.2 Consequentialism 20.2.3 Deontological Ethics 20.2.4 Virtue Ethics 20.2.5 The Focus of Academic Data Ethics 20.3 Data Ethics in the Commercial Domain 20.3.1 Technological Level 20.3.2 Individual Level 20.3.3 Organizational Level 20.4 Law and Data Ethics 20.5 Data Ethics and Data Science: Are They in It for the Long Run? References 21: Value-Sensitive Software Design 21.1 Introduction 21.2 The Good, the Bad, and the Never Neutral 21.2.1 Non-neutrality 21.2.2 Impact on a Micro-level 21.2.3 Impact on a Macro-level 21.2.4 In Sum 21.3 Employing the Never Neutral 21.3.1 A Challenge for Designers 21.3.2 Value-Sensitive Design 21.3.3 Values 21.3.4 Legal Values and Design References 22: Data Science for Entrepreneurship: The Road Ahead 22.1 Introduction 22.2 The Road Ahead 22.2.1 AI Software 22.2.2 MLOps 22.2.3 Edge Computing 22.2.4 Digital Twins 22.2.5 Large-Scale Experimentation 22.2.6 Big Data and AI Opportunities 22.2.7 Government Regulation References