Federal Data Science

Description

Federal Data Science serves as a guide for federal software engineers, government analysts, economists, researchers, data scientists, and engineering managers in deploying data analytics methods to governmental processes. Driven by open government (2009) and big data (2012) initiatives, federal agencies have a serious need to implement intelligent data management methods, share their data, and deploy advanced analytics to their processes. Using federal data for reactive decision making is not sufficient anymore, intelligent data systems allow for proactive activities that lead to benefits such as: improved citizen services, higher accountability, reduced delivery inefficiencies, lower costs, enhanced national insights, and better policy making.

No other government-dedicated work has been found in literature that addresses this broad topic. This book provides multiple use-cases, describes federal data science benefits, and fills the gap in this critical and timely area. Written and reviewed by academics, industry experts, and federal analysts, the problems and challenges of developing data systems for government agencies is presented by actual developers, designers, and users of those systems, providing a unique and valuable real-world perspective.

Offers a range of data science models, engineering tools, and federal use-cases
Provides foundational observations into government data resources and requirements
Introduces experiences and examples of data openness from the US and other countries
A step-by-step guide for the conversion of government towards data-driven policy making
Focuses on presenting data models that work within the constraints of the US government
Presents the why, the what, and the how of injecting AI into federal culture and software systems

Front Cover 2
Federal Data Science 3
Federal Data Science:Transforming Government and Agricultural Policy Using Artificial Intelligence 5
Copyright 6
Dedication 7
Contents 9
List of Contributors 15
About the Editors 17
Note From the Editors 19
Foreword 21
Preface 25
1 - 1Injecting Artificial Intelligence Into Governmental Systems 29
- 1 - A Day in the Life of a Federal Analyst and a Federal Contractor 31
  - 1. IN THE EARLY MORNING 31
  - 2. LATER IN THE AFTERNOON 33

Front Cover 2
Federal Data Science 3
Federal Data Science:Transforming Government and Agricultural Policy Using Artificial Intelligence 5
Copyright 6
Dedication 7
Contents 9
List of Contributors 15
About the Editors 17
Note From the Editors 19
Foreword 21
Preface 25
1 - 1Injecting Artificial Intelligence Into Governmental Systems 29
- 1 - A Day in the Life of a Federal Analyst and a Federal Contractor 31
  - 1. IN THE EARLY MORNING 31
  - 2. LATER IN THE AFTERNOON 33
  - 3. LATE, LATE AT NIGHT 36
  - 4. THEREFORE, THIS BOOK 38
  - REFERENCES 40
- 2 - Disseminating Government Data Effectively in the Age of Open Data 41
  - 1. DATA DISSEMINATION IN THE FEDERAL GOVERNMENT: FROM COLONIAL AMERICA TO OPEN DATA 41
  - 2. OPEN DATA POLICY: A NEW ERA IN DATA DISSEMINATION 42
  - 3. NEW ERA, NEW CHALLENGES 44
  - 4. TOWARD A MODERN AND EFFECTIVE DATA DISSEMINATION STRATEGY 51
  - 5. CONCLUSION 55
  - DISCLAIMER 55
  - REFERENCES 56
- 3 - Machine Learning for the Government: Challenges and Statistical Difficulties 57
  - 1. INTRODUCTION 57
  - 2. AN INTRODUCTION TO DATA MINING 58
    - 2.1 Learning With Orange 59
  - 3. RESULT VALIDATION, TRUST BUT VERIFY 62
    - 3.1 Iris Aside 65
  - 4. MODEL OVERFITTING, TOO GOOD TO BE TRUE 65
  - 5. STATISTICAL BIAS, IMPACTING RESULTS BEFORE ANALYSIS BEGINS 65
  - 6. SEGMENTATION AND SIMPSON’S PARADOX 66
  - 7. OUTLIERS AND BAD DATA 66
  - 9. CONCLUSION 67
  - 8. NONREPRODUCIBILITY AND STATISTICS HUNTING 67
  - REFERENCES 68
- 4 - Making the Case for Artificial Intelligence at Government: Guidelines to Transforming Federal Software Systems 69
  - 1. MOTIVATIONS AND OBJECTIONS 69
  - 2. AI TECHNOLOGIES AND GOVERNMENT USE CASES 72
    - 2.1 Knowledge-Based Systems 72
    - 2.2 Big Data 73
    - 2.3 Machine Learning and Data Analytics 74
  - 3. CONCLUSIONS 76
  - REFERENCES 78
- 2 - Governmental Data Science SolutionsAround the World 81
  - 5 - Agricultural Data Analytics for Environmental Monitoring in Canada 83
    - 1. INTRODUCTION 83
    - 2. MATERIALS AND METHODS 86
      - 2.1 Input Data 86
      - 2.2 Input Preparation 87
      - 2.5 Rule Development 88
      - 2.3 Input Accuracy Assessment 88
      - 2.4 Coregistration 88
      - 2.6 Contextual Assessment and Rectification 93
      - 2.7 Assessment of Class Distributions 96
      - 2.8 Output Accuracy Assessment 97
    - 3. RESULTS AND DISCUSSION 97
      - 3.1 Accuracy Assessment 97
      - 3.2 National-Scale Maps 102
      - 3.3 Land Use Change 104
    - 4. CONCLUSIONS 105
    - ACKNOWLEDGMENTS 106
    - REFERENCES 106
  - 6 - France’s Governmental Big Data Analytics: From Predictive to Prescriptive Using R 109
    - 1. INTRODUCTION 109
    - 2. MATERIALS AND METHODS: “PARSIMONIOUS MODELING FOR PRESCRIPTIVE DATA SCIENCE,” APPLIED TO FRENCH AGRICULTURE 110
      - 2.1 Agricultural Data in France 110
      - 2.2 Open Taxonomies for Agricultural Data Sciences 111
      - 2.3 Big Data and Data Science in France 112
    - 3. RESULTS 112
      - 3.1 A Parsimonious Agricultural Taxonomy for Data Collection, an Intermediate Result 112
      - 3.2 Agricultural Descriptive and Predictive Data Science With R 115
        
        3.2.3 Prediction of Soil Fertility 116
        
        3.2.1 Crop Monitoring 116
        
        3.2.2 Evapotranspiration 116
      - 3.3 From Descriptive Analytics to Prescriptive Analytics Through Predictive Analytics 116
        
        3.3.1 Toward Prescriptive Analytics 116
        
        3.3.2 Three Main Kinds of Aggregation Engine 117
        
        3.3.2.2 Optimization of a Utility Function 118
        
        3.3.2.3 Heuristic and Fuzzy 118
        
        3.3.2.1 Ensemble Methods 118
        
        3.3.3 Available R Building Blocks for Building an Aggregation Engine (Nonexhaustive Mapping) 119
      - 4. CONCLUSION 119
      - REFERENCES 120
    - 7 - Agricultural Remote Sensing and Data Science in China 123
      - 1. AGRICULTURAL REMOTE SENSING IN CHINA 123
        
        1.1 Agricultural Remote Sensing Research and Applications 123
        
        1.2 China Agricultural Remote Sensing Monitoring System 127
      - 2. DATA SCIENCE IN CHINA 127
        
        2.1 Data Science Development in China 127
        
        2.2 Science Data Sharing and Services Platforms 131
      - 3. CONCLUSIONS 133
      - ACKNOWLEDGMENTS 134
      - REFERENCES 134
    - 8 - Data Visualization of Complex Information Through Mind Mapping in Spain and the European Union 137
      - 1. DATA SCIENCE ECOSYSTEM IN THE EUROPEAN UNION 137
        
        1.1 Horizon 2020 (Horizon 2020 Documents, 2016) 137
        
        1.3 Open Data Incubator Europe 138
        
        1.2 The European Data Landscape Monitoring Tool (European Data Landscape Monitoring Tool Documents, 2017) 138
        
        1.4 Data Science Education in the European Union 139
        
        1.4.1 European Data Science Academy 139
        
        1.4.2 Educational Curriculum for the Usage of Linked Data (EUCLID Documents, 2017) 140
        
        1.4.3 Master Course in Data Mining and Knowledge Management 140
        
        1.4.4 European Institute of Innovation & Technology (EIT Documents, 2017) Digital Master School. Data Science Master (EIT Data S 141
        
        1.4.5 Erasmus Mundus Joint Master Degree Program in Big Data Management and Analytics (BDMA, 2017) 141
        
        1.5 Other Organizations 142
        
        1.5.1 Open Data Institute 142
        
        1.5.2 European Data Forum (EDF Documents, 2016) 142
        
        1.5.3 Big Data Value Association (BDV Documents, 2017) 143
        
        1.6 Data Science and Big Data in Spain 143
      - 2. OPEN DATA IN THE EUROPEAN UNION AND SPAIN 144
        
        2.1 Open Data in the European Union 144
        
        2.2 Open Data in Spain 145
      - 3. VISUALIZATION OF BIG DATA AND OPEN DATA 145
      - 4. MIND MAPPING 146
        
        4.1 Introduction 146
        
        4.2 Digital Mind Maps 147
        
        4.3 The Importance of Mind Mapping 148
        
        4.4 Advantages of Mind Mapping 149
        
        4.5 Experiments and Surveys Related to Mind Mapping 150
        
        4.6 Use of Mind Mapping in Governments 152
        
        4.7 Mind Mapping Automation 153
        
        4.7.3 Open Data (Infoseg, 2013c,d) 154
        
        4.7.1 Statistical Analysis With R (Infoseg Documents, 2013a,b) 154
        
        4.7.2 Big Open Data Analysis Using R (Infoseg, 2014a,b) 154
        
        4.7.6 Mind Mapping Integration in Artificial Intelligence Systems 155
        
        4.7.4 Complex Information Systems 155
        
        4.7.5 Analysis of Information Security Log Files 155
        
        5. USES OF MIND MAPPING IN THE FEDERAL GOVERNMENT 156
        
        6. CONCLUSIONS 159
        
        REFERENCES 160
      - 3 - Federal Data ScienceUse Cases at the USGovernment 167
        
        9 - A Deployment Life Cycle Model for Agricultural Data Systems Using Kansei Engineering and Association Rules 169
        
        1. INTRODUCTION AND BACKGROUND 169
        
        1.2 Motivation 170
        
        1.1 A Measuring Stick 170
        
        1.3 Systems Life Cycle Models 171
        
        1.4 Analytical Models for the Government 172
        
        2. RELATED WORK 174
        
        2.1 Intelligent Software Testing 174
        
        2.2 Kansei and Software Deployment (A Review) 178
        
        3. THE FEDERAL DEPLOYMENT AND ADOPTION LIFE CYCLE 179
        
        3.1 Association Rules Testing 179
        
        3.2 Kansei Engineering Deployment and Traceability 180
        
        4. EXPERIMENTAL STUDIES ON KANSEI ENGINEERING DEPLOYMENT AND TRACEABILITY AND ASSOCIATION RULES TESTING 182
        
        4.1 Code Coverage and Maintenance Costs of Association Rules Testing 183
        
        4.2 The Agricultural Analyst’s Kansei Survey 184
        
        5. CONCLUSIONS AND FUTURE WORK 185
        
        REFERENCES 186
        
        FURTHER READING 187
        
        10 - Federal Big Data Analytics in the Health Domain: An Ontological Approach to Data Interoperability 189
        
        1. INTRODUCTION 189
        
        2. DATA INTEROPERABILITY IN THE HEALTH DOMAIN 190
        
        3. ONTOLOGIES AS THE BASIS FOR INTEROPERABILITY 194
        
        3.1 Lexicon as the Basis for Semantic Congruity 194
        
        3.2 Ontological Dimensions 194
        
        3.3 Ontology Development 196
        
        3.4 Ontology Integration 199
        
        3.5 Ontology Operationalization 200
        
        3.6 Metadata Foundations 201
        
        4. CONCLUSION 201
        
        REFERENCES 202
        
        11 - Geospatial Data Discovery, Management, and Analysis at National Aeronautics and Space Administration (NASA) 205
        
        1. INTRODUCTION 205
        
        2. GEOSPATIAL DATA DISCOVERY 206
        
        3. BIG GEOSPATIAL DATA MANAGEMENT 208
        
        4. LARGE-SCALE SCIENTIFIC SIMULATION 209
        
        4.1 Spatiotemporal Thinking to Optimize High-Performance Computing 210
        
        4.2 Cloud Computing to Support Large-Scale Scientific Simulation 212
        
        5. SPATIOTEMPORAL DATA MODELING AND ANALYSIS 212
        
        5.1 Spatiotemporal Data Model 213
        
        5.2 Tracking Changes and Interactions 213
        
        5.3 Spatiotemporal Analytics 215
        
        6. CONCLUSION AND FUTURE DIRECTIONS 216
        
        REFERENCES 217
        
        ACKNOWLEDGMENTS 217
        
        12 - Intelligent Automation Tools and Software Engines for Managing Federal Agricultural Data 221
        
        1. INTRODUCTION AND MOTIVATION 221
        
        2. RELATED WORK 222
        
        2.1 Data Validation Methods 222
        
        2.2 Data Security and Integrity Methods 223
        
        3. THE INTELLIGENT FEDERAL MATH ENGINE 224
        
        3.1 Inputs, Outputs, and Process of the Math Engine 224
        
        3.2 The Seven-Step Math Process 226
        
        4. VALIDATION AND VERIFICATION OF FEDERAL AGRICULTURAL DATA 227
        
        5. THE INTELLIGENT FEDERAL DATA MANAGEMENT TOOL 230
        
        5.1 Federal Tool Requirements 230
        
        5.2 Federal Tool Implementation 231
        
        6. INSIGHTS, EXPERIMENTAL WORK, AND CONCLUSIONS 233
        
        6.1 Experimental Setup and Results 233
        
        6.2 Lessons Learnt and Keys to Technical Federal Success 235
        
        REFERENCES 237
        
        FURTHER READING 238
        
        13 - Transforming Governmental Data Science Teams in the Future 239
        
        1. INTRODUCTION 239
        
        2. SITUATIONAL LEADERSHIP 240
        
        3. ARCHETYPES 243
        
        3.1 Mapping Archetypes to Prior Career Paths 243
        
        3.2 Archetypes in the Federal Government: An Example 244
        
        3.3 Archetypes, Risk Mitigation, and Growth 245
        
        4. BEST PRACTICES 246
        
        4.1 Creation of a Best Practice 246
        
        4.2 Intra- and Interagency Collaboration 247
        
        5. CONCLUSION 248
        
        ACKNOWLEDGMENT 249
        
        REFERENCES 249
        
        Afterword 251
        
        Index 253
        
        A 253
        
        B 253
        
        E 254
        
        C 254
        
        D 254
        
        H 255
        
        L 255
        
        F 255
        
        I 255
        
        G 255
        
        K 255
        
        N 256
        
        R 256
        
        O 256
        
        S 256
        
        M 256
        
        P 256
        
        U 257
        
        V 257
        
        T 257
        
        Back Cover 259

Transforming Government and Agricultural Policy Using Artificial Intelligence

Description

TOC

Reviews (0)