Artificial Intelligence for .NET: Speech, Language, and Search: Building Smart Applications with Microsoft Cognitive Services APIs

Contents at a Glance......Page 5
Contents......Page 6
About the Author......Page 14
About the Contributing Author......Page 15
About the Technical Reviewer......Page 16
Acknowledgments......Page 17
Introduction......Page 18
Chapter 1: Getting Started with AI Basics......Page 20
 Truth vs. Fiction......Page 21
 History and Evolution......Page 22
  Commoditization of AI......Page 23
 Microsoft and AI......Page 24
 Basic Concepts......Page 25
  Machine Learning......Page 28
   Types of Machine Learning......Page 30
   Natural Language Processing......Page 31
   TTS and STT......Page 32
   Object Detection......Page 33
   Image Recognition......Page 34
   Optical Character Recognition......Page 35
 Microsoft’s Cognitive Services......Page 36
  Speech......Page 37
  Knowledge......Page 38
  Search......Page 39
 Recap......Page 40
Chapter 2: Creating an AI-Based Application in Visual Studio......Page 41
 Getting an Azure Subscription Key for Cognitive Services......Page 42
  Step 1: Set Up an Azure Account......Page 43
  Step 2: Create a New Cognitive Services Account......Page 45
  Step 3: Get the Subscription Key(s)......Page 47
 Testing the API......Page 48
  How To Do It......Page 49
 Creating Your First AI-based Application......Page 51
  The Code......Page 52
  The Walkthrough......Page 54
  The Result......Page 56
  The Code......Page 57
  The Walkthrough......Page 59
  The Result......Page 60
 Recap......Page 61
Chapter 3: Building a Conversational User Interface with Microsoft Technologies......Page 62
  In the Very Beginning: the Command-Line Interface (CLI)......Page 64
  And Then Came the Graphical User Interface......Page 66
  And UI Evolved Yet Again: Conversational User Interface......Page 68
  AI’s Role in CUI......Page 69
  Pitfalls of CUI......Page 70
  A Hybrid UI (CUI+GUI) May Be the Future......Page 72
 Design Principles......Page 74
 Creating a CUI Application Using Bot Framework......Page 75
  Step 0: Prep Your Development Environment......Page 76
  Step 2: First and Default Messages......Page 77
  Step 3: Running, Testing, and Debugging Your Bot......Page 79
  Step 3: Appointment Scheduling......Page 82
  Next Steps......Page 85
 Recap......Page 86
Chapter 4: Using Natural Language Understanding......Page 87
 What Is NLU?......Page 88
 History of Natural Language Understanding......Page 90
  Complexities in Natural Language......Page 93
  Statistical Models as a Solution Are Insufficient......Page 95
 Language Understanding Intelligent Service (LUIS)......Page 96
  Architecture of a LUIS-Based Software Application......Page 97
  Behind the Scenes......Page 100
  Extensive Training Is the Key......Page 101
 Getting an Azure Subscription for LUIS......Page 102
  Getting Subscription Keys on Azure Portal......Page 103
  Applying Subscription Keys in LUIS......Page 104
 Demo: Definition App......Page 105
 Recap......Page 108
 The Bing Spell Check API......Page 109
  What Is It?......Page 111
   Request......Page 112
   Response......Page 114
  Integration with LUIS......Page 115
 The Text Analytics API......Page 117
  Language Detection......Page 118
   Request......Page 119
   Response......Page 120
  Key Phrase Extraction......Page 121
   Request......Page 123
  Sentiment Analysis......Page 124
   Request......Page 125
  Topic Detection......Page 126
   Request......Page 127
   Response......Page 128
  Usage Ideas......Page 129
 The Web Language Model (WebLM) API......Page 130
   Response......Page 132
   Request......Page 133
   Response......Page 134
   Request......Page 135
   Request......Page 136
 The Linguistic Analysis API......Page 137
  Sentence Separation and Tokenization......Page 138
   Response......Page 139
  Part-of-Speech Tagging......Page 141
   Request......Page 142
  Constituency Parsing......Page 143
   Response......Page 145
 Recap......Page 146
 Planning Your App......Page 147
  What Should Be Done in LUIS?......Page 148
 Creating a LUIS App......Page 150
 Adding/Labeling Utterances......Page 151
 Publishing Your App......Page 153
 Adding Entities......Page 155
  Simple Entities......Page 157
  Composite Entities......Page 158
  Hierachical Entities......Page 161
  Prebuilt Entities......Page 163
 Adding a Phrase List......Page 165
  Active Learning Through Suggested Utterances......Page 166
  Creating a Project in Visual Studio......Page 167
  Handling an Entity-less Intent......Page 168
  Testing the Bot in an Emulator......Page 169
  Handling an Entity-Full Intent......Page 170
  Handling an Intent with Composite Entities......Page 172
  Publishing Your Bot......Page 174
  Registering Your Bot......Page 175
 Recap......Page 176
Chapter 7: Interacting with the Speech API......Page 177
 Ways to Interact with Speech......Page 178
  The Cognitive Search API......Page 179
 Getting the JSON Web Token First......Page 180
 The Consume Speech API......Page 182
 Speech Synthesis......Page 183
  Speech Recognition Internals......Page 186
  Custom Acoustic Model......Page 187
  Custom Language Model......Page 196
  Pronunciation Data......Page 198
  Custom Speech-to-Text Endpoint......Page 199
 Speaker Recognition......Page 201
  Enrollment-Verification......Page 202
  Speaker Verification......Page 205
  Enrollment–Identification......Page 206
  Operation Status......Page 207
 Summary......Page 208
 Search Is Everywhere......Page 209
 Pervasive, Predictive, Proactive (The Three Ps of Search)......Page 211
 History of Bing......Page 212
 Search APIs......Page 213
  Bing Autosuggest API......Page 214
  How to Consume the Bing Autosuggest API......Page 215
 The Bing Image Search API......Page 218
  How to Consume the Bing Image Search API......Page 219
 Bing News Search API......Page 224
 Bing Video Search API......Page 227
 How to Consume the Bing Video Search API......Page 228
 Bing Web Search API......Page 231
  How to Consume the Bing Web Search API......Page 232
 Summary......Page 235
Chapter 9: Working with Recommendations......Page 236
 Understanding the Basics......Page 237
 Frequent Brought Together (FBT) Recommendations......Page 238
 Recommendations Based on Past History......Page 239
 How Do These Recommendations Work?......Page 240
 Recommendation Models and Types......Page 244
 Recommendation Build......Page 245
 Frequent Brought Together (FBT) Build......Page 249
 Ranking Recommendation......Page 251
 SAR (Smart Adaptive Recommendations) Build......Page 253
 Setting Rules in Build......Page 255
 Offline Evaluation......Page 256
 Recommendation UI......Page 257
 Summary......Page 261
 Why Is AI So Popular?......Page 262
 Improved Computing Power......Page 263
 Data Is the New Currency......Page 264
 Services vs Solutions?......Page 266
 Challenges and the Future of NLU......Page 267
 Challenges and the Future of Search......Page 268
 Challenges and the Future of Recommendations......Page 269
 Intelligent Edge......Page 270
 Tasks, not Jobs, Will Be Eliminated......Page 271
 So Where Do We Go From Here?......Page 273
Index......Page 275