11 - QnA Bot.ipynb

{"cells":[{"cell_type":"markdown","source":["# Conversational AI\r\n","\r\n","Think about how often you communicate with other people through instant messaging, social media, email, or other online technologies. For many of us, it's our go-to form of contact. When you have a question at work, you might reach out to a colleague using a chat message, which you can use on mobile devices, so you're always in touch.\r\n","\r\n","![A human and a robot having a conversation](./images/conversational_ai.jpg)\r\n","\r\n","Bots are AI agents that communicate using these kinds of channels, enabling a natural, conversational engagement with software services.\r\n","\r\n","## Create a QnA Maker Knowledge Base\r\n","\r\n","For customer support scenarios, it's common to create a bot that can interpret and answer frequently asked questions through a website chat window, email, or voice interface. Underlying the bot interface is a knowledge base of questions and appropriate answers that the bot can search for suitable responses.\r\n","\r\n","The QnA Maker service is a cognitive service in Azure that enables you to quickly create a knowledge base, either by entering question and answer pairs or from an existing document or web page. It can then use some built-in natural language processing capabilities to interpret questions and find appropriate answers.\r\n","\r\n","1. Open another browser tab and go to the QnA Maker portal at https://qnamaker.ai. Sign in using the Microsoft account associated with your Azure subscription.\r\n","2. In the QnA Maker portal, select **Create a knowledge base**.\r\n","3. If you haven't previously created a QnA service resource, select **Create a QnA service**. The Azure portal will be opened in another tab so you can create a QnA Maker service in your subscription. Use the following settings:\r\n","   \r\n","   - **Subscription**: *Your Azure subscription*\r\n","   - **Resource group**: *Select an existing resource group or create a new one*\r\n","   - **Name**: *A unique name for your QnA resource*\r\n","   - **Pricing tier**: F0\r\n","   - **Azure Search pricing tier**: F\r\n","   - **Azure Search location**: *Any available location*\r\n","   - **App name**: _Same as **Name** (\".azurewebsites.net\" will be appended automatically)_\r\n","   - **Website location**: _Same as **Azure Search location**_\r\n","   - **App insights**: Disable\r\n","\r\n","> **Note**: If you have already provisioned a free-tier **QnA Maker** or **Azure Search** resources, your quota may not allow you to create another one. In which case, select a tier other than **F0** / **F**.\r\n","\r\n","4. Wait for the deployment of the QnA Service and related resources to complete in the Azure portal.\r\n","5. Return to the QnA Maker portal tab, and in the **Step 2** section, click **Refresh** to refresh the list of available QnA service resources.\r\n","6. Connect your QnA service to your KB by selecting the following options:\r\n","  - **Microsoft Azure Directory ID**: *The Azure directory ID for your subscription*\r\n","  - **Azure subscription name**: *Your Azure subscription*\r\n","  - **Azure QnA service**: *The QnA service resource you created in the previous step*\r\n","  - **Language**: English\r\n","\r\n","### (!) Check In\r\n","If a message stating that the role does not have permission to perform the action is displayed, refresh the browser page for the QnA Maker portal.\r\n","\r\n","7. In the **Step 3** section, enter the name **Margie's Travel KB**.\r\n","8. In the **Step 4** section, in the **URL** box, type *https://github.com/MicrosoftLearning/mslearn-ai900/raw/main/data/qna_bot/margies_faq.docx* and click **+ Add URL**. Then under **Chit-chat**, select **Professional**.\r\n","9. In the **Step 5** section, click **Create your KB**.\r\n","10. Wait for a minute or so while your Knowledge base is created. Then review the questions and answers that have been imported from the FAQ document and the professional chit-chat pre-defined responses.\r\n"],"metadata":{}},{"cell_type":"markdown","source":["## Edit the Knowledge Base\n","\n","Your knowledge base is based on the details in the FAQ document and some pre-defined responses. You can add custom question-and-answer pairs to supplement these.\n","\n","1. Click **+ Add QnA pair**.\n","2. In the **Question** box, type `Hello`. Then click **+ Add alternative phrasing** and type `Hi`.\n","3. In the **Answer** box, type `Hello`.\n","\n","## Train and Test the Knowledge Base\n","\n","Now that you have a knowledge base, you can test it in the QnA Maker portal.\n","\n","1. At the top right of the page, click **Save and train** to train your knowledge base. You may need to change the size of your window to see the button.\n","2. After training has completed, click **&larr; Test** to open the test pane.\n","3. In the test pane, at the bottom enter the message *Hi*. The response **Hello** should be returned.\n","4. In the test pane, at the bottom enter the message *I want to book a flight*. An appropriate response from the FAQ should be returned.\n","5. When you're done testing the knowledge base, click **&rarr; Test** to close the test pane.\n"],"metadata":{}},{"cell_type":"markdown","source":["## Create a Bot for the Knowledge Base\n","\n","The knowledge base provides a back-end service that client applications can use to answer questions through some sort of user interface. Commonly, these client applications are bots. To make the knowledge base available to a bot, you must publish it as a service that can be accessed over HTTP. You can then use the Azure Bot Service to create and host a bot that uses the knowledge base to answer user questions.\n","\n","1. At the top of the QnA Make page, click **Publish**. Then in the **Margies Travel KB** page, click **Publish**.\n","2. After the service has been deployed, click **Create Bot**. This opens the Azure portal in a new browser tab so you can create a Web App Bot in your Azure subscription.\n","3. In the Azure portal, create a Web App Bot with the following settings (most of these will be pre-populated for you):\n","  - **Bot handle**: *A unique name for your bot*\n","  - **Subscription**: *Your Azure subscription*\n","  - **Resource group**: *The resource group containing your QnA Maker resource*\n","  - **Location**: *The same location as your QnA Maker service*.\n","  - **Pricing tier**: F0\n","  - **App name**: *Same as the **Bot handle** with *.azurewebsites.net* appended automatically\n","  - **SDK language**: *Choose either C# or Node.js*\n","  - **QnA Auth Key**: *This should automatically be set to the authentication key for your QnA knowledge base*\n","  - **App service plan/location**: *This should be set automatically to a suitable plan and location*\n","  - **Application Insights**: Off\n","  - **Microsoft App ID and password**: Auto create App ID and password.\n","4. Wait for your bot to be created (the notification icon at the top right, which looks like a bell, will be animated while you wait). Then in the notification that deployment has completed, click **Go to resource** (or alternatively, on the home page, click **Resource groups*, open the resource group where you created the web app bot, and click it.)\n","5. In the left-hand pane of your bot look for **Settings**, click on **Test in Web Chat**, and wait until the bot displays the message **Hello and welcome!** (it may take a few seconds to initialize).\n","6. Use the test chat interface to ensure your bot answers questions from your knowledge base as expected. For example, try submitting *I need to cancel my hotel*.\n"],"metadata":{}},{"cell_type":"markdown","source":["## Access the Bot through a Channel\n","\n","A bot can be used to provide an interface for users through one or more *channels*. For example, the same bot could support interactions through a web chat interface, email, and Microsoft Teams.\n","\n","1. In the left-hand pane of your bot look for **Settings** and click on **Channels**.\n","2. Note that the **Web Chat** channel has been added automatically, and that other channels for common communication platforms are available.\n","3. Next to the **Web Chat** channel, click **Edit**. This opens a page with the settings you need to embed your bot in a web page. To embed your bot, you need the HTML embed code provided as well as one of the secret keys generated for your bot.\n","4. Copy the **Embed code** and paste it into the cell below, replacing the comment `<!-- EMBED CODE GOES HERE -->`.\n","5. Click **Show** for one of your secret keys (it doesn't matter which one), and copy it. Then paste it in your HTML embed code below, replacing `YOUR_SECRET_HERE`.\n","6. Change the **min-height** value in your HTML code to **200px** (instead of the default 500px). This will help ensure the HTML interface is visible without scrolling.\n","7. Run the cell below by clicking the **Run cell** (&#9655;) button on the left of the cell to render the HTML.\n","8. In the HTML chat interface, test the bot by submitting a question, such as *Who is Margie?* or *What destinations can I go to?* (when the bot initializes, it may respond with the message *Hello and welcome* in addition to answering your question.)."],"metadata":{}},{"cell_type":"code","execution_count":null,"source":["%%html\r\n","\r\n","<!-- EMBED CODE GOES HERE -->"],"outputs":[],"metadata":{}},{"cell_type":"markdown","source":["Experiment with the bot. You'll probably find that it can answer questions from the FAQ quite accurately, but it will have limited ability to interpret questions that it has not been trained with. You can always use the QnA Maker portal to edit the knowledge base to improve it, and republish it.\n","\n","## Learn More\n","\n","- To learn more about the QnA Maker service, view [the QnA Maker documentation](https://docs.microsoft.com/azure/cognitive-services/qnamaker/).\n","- To learn more about the Microsoft Bot Service, view [the Azure Bot Service page](https://azure.microsoft.com/services/bot-service/)."],"metadata":{}}],"metadata":{"language_info":{"name":"python","version":"3.6.9","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kernelspec":{"name":"python3-azureml","language":"python","display_name":"Python 3.6 - AzureML"},"kernel_info":{"name":"python3-azureml"},"nteract":{"version":"nteract-front-end@1.0.0"}},"nbformat":4,"nbformat_minor":2}