Getting Started With QnA Maker Service

What is QnA Maker service?

In simple language, it is a service to educate a bot with static answers for a specified set of questions. Let’s take an example of FAQs. FAQs are sets of questions and answers which are going to remain the same irrespective of the user.

Now, we will jump into the configuration part of service.

• To start with QnA Maker service, go to qnamaker.ai
• Click on sign in and log in with your Microsoft account. You can see 'My Services' and 'Create new service' tab. 'My services' tab will list down all QnA services you have created earlier.
  
  
  

To create new instance of service, click on 'Create new service' tab. It will redirect you to following page.



Here you can specify name for service. I have named it as LUIS FAQs Service. As I am going to use this for FAQs related to LUIS.

There are three options to educate QnA maker for your set of questions and answers.

Option 1

If you have a webpage with FAQs content, you can specify URL of that page here. QnA maker will consume all QnAs from specified resources. I have specified URL of LUIS resources FAQ.

Option 2

You can also upload a file with questions and answers with specified format. You can download example file by clicking on 'See an example link'.

Option 3

You can specify your QnAs from scratch by adding them one by one. But before that you need to click on Create link.

When you click on Create link, it will redirect to Knowledge Base page. Here you can see QnAs you have specified on last page of configuration with the help of option 1 or option 2.

For option 3, you can click on 'Add new QnA pair' link in right upper corner.



After specifying QnA pairs, now its time to test our service. Click on Test link below Knowledge Base link.



QnA maker provide us chat emulator to test our service. I have asked question regarding 'managing spelling mistakes' in LUIS.

QnA maker does following steps while searching for answer,

1. NLP (natural language processing) for a question I have asked.
2. Match with existing set of questions and rank them as per relevance.
3. Answer of question with highest rank will be returned by service.

Click on 'Settings' link for more options. Under settings you can share the service with other Microsoft accounts by specifying their email ids.

'Save and retrain'

It saves all settings and re-educate QnA maker for any changes made.

'Publish'

After publishing, service can be accessible over HTTP. You can get an answer for question by HTTP request against service.



Before publishing, it will show changes for review. Here you can see sources for QnAs. You can also download diff files to learn the difference between subsequent publishes.



After clicking on publish, it will show success message and sample HTTP request. You can use this to query QnA maker service through bot application using Microsoft Bot Framework. In the next post, we will look into how to get answers from QnA maker service into bot applications.

In next article FAQs Answering bot using Bot Framework and QnA Maker service, we will create FAQs Answering bot using Microsoft bot framework and QnA Maker service created in this article.

Read more on Microsoft Bot Framework 

• Getting Started With QnA Maker Service
• Quick start - Development Of Chat Bot Using Microsoft Bot Framework - For Beginners
• FAQs Answering bot using Bot Framework and QnA Maker service
• Chat Bot Using Microsoft Bot Framework With LUIS - Part One
• Chat Bot Using Microsoft Bot Framework With LUIS - Part Two
• Chat Bot With Azure Authentication - Part One
• Chat Bot With Azure Authentication - Part Two
• Chat Bot with SharePoint authentication - Part One
• Chat Bot with SharePoint authentication - Part Two
• Develop ChatBot On NodeJS platform Using Microsoft Bot Framework (Part One) - Quick Start For Beginners
• Develop ChatBot On NodeJS Platform Using Microsoft Bot Framework (Part Two) - Manage Conversation Using Root Dialog
• Develop Chatbot On NodeJS Platform Using Microsoft Bot Framework (Part Three) - Manage Conversation Using Multiple Dialogs