Book Image

Learning Microsoft Cognitive Services - Third Edition

By : Leif Larsen
Book Image

Learning Microsoft Cognitive Services - Third Edition

By: Leif Larsen

Overview of this book

Microsoft Cognitive Services is a set of APIs for integrating artificial intelligence in your applications to solve logical business problems. If you’re new to developing applications with AI, Learning Microsoft Cognitive Services will give you a comprehensive introduction to Microsoft’s AI stack and get you up-to-speed in no time. The book introduces you to 24 APIs, including Emotion, Language, Vision, Speech, Knowledge, and Search. Using Visual Studio, you can develop applications with enhanced capabilities for image processing, speech recognition, text processing, and much more. Moving forward, you will work with datasets that enable your applications to process various data in the form of image, video, or text. By the end of the book, you’ll be able to confidently explore Cognitive Services APIs for building intelligent applications that can be deployed for real-world business uses.
Table of Contents (17 chapters)
Learning Microsoft Cognitive Services - Third Edition
Contributors
Acknowledgments
Preface
Index

Training the model


Every time we make changes to the knowledge base, it is wise to click Save and Train. This will ensure that our model is up to date, with the most current question-and-answer pairs.

Once we have trained the model, we can test it. This can be done by clicking the Test button on the right-hand side. This will present us with the following chat window:

From this chat dialog, we can test some or all of our questions to verify that we get the correct answers. We can also improve the model by asking questions in different ways. In some cases, this will present us with the wrong answer.

If we have been presented with the wrong answer, we can change this by selecting the correct one. With any given question, the possible answers will be listed by clicking the Inspect button beneath the question, ordered by probability. Selecting the correct answer and retraining the model will ensure a correct answer when asking the same question later.