Determine the Sentiments behind interaction between Users and our Bot with Azure Cognitive Services

In Artificial Intelligence by Christian HissibiniLeave a Comment


The interaction between users and bots is mostly free-form, so bots need to understand language naturally and contextually. In this exercise you will learn how to detect the user’s sentiments and mood using the Azure Text Analytics API.

With Text Analytics APIs, part of the Azure Cognitive Services offering, you can detect sentiment, key phrases, topics, and language from your text. The API returns a numeric score between 0 and 1. Scores close to 1 indicate positive sentiment and scores close to 0 indicate negative sentiment. Sentiment score is generated using classification techniques.

Inside this folder you will find a solution with the code that results from completing the steps in this exercise. You can use this solution as guidance if you need additional help as you work through this exercise. Remember that before using it, you first need to complete the values in Web.config.


The following software is required for completing this exercise:

Task 1: Create the Text Analytics API Key

In this task you will create a Text Analytics account.

  1. Browse here, select the Language tab. Find the Text Analytics API and click Create. You will be prompted to agree the terms of use and choose your country, next click Next.
  2. Log in with your Azure Subscription account. You should be taken to a page like the following one with an evaluation key with 5000 free requests per month. Save one of the keys for later.

Task 2: Add the Text Analytics API Client

In this task you will create a new class to call the Text Analytics API from the bot.

  1. Open the solution you’ve obtained from exercise 4. Alternatively, you can use the app from the exercise4-KnowledgeBase folder.NOTE: If you use the solution provided remember to replace:
    • the [LuisModel(“{LUISAppID}”, “{LUISKey}”)] attribute in Dialogs\RootDialog.cs with your LUIS App Id and Programmatic API Key (as explained in exercise 3)
    • the AzureSearchAccount and AzureSearchKey in Web.config with your search account name and key (as explained in exercise 4)
  2. Copy the TextAnalyticsService.cs in the project’s Services folder. This file contains three classes to consume the Text Analytics API.NOTE: Notice that the client is hitting the /sentiment endpoint. The Text Analytics API also provides the /keyPhrases and /languages endpoints. Also notice that you can send more than one document to analyze.
  3. Update your Web.Config file in your project’s root folder adding the key TextAnalyticsApiKey under the appSettingssection. Complete the TextAnalyticsApiKey value with the Text Analytics key you’ve obtained in the previous task.<add key=”TextAnalyticsApiKey” value=”” />
  4. In the Dialogs folder, create a new class UserFeedbackRequestDialog.cs using the following boilerplate code. This dialog will have the responsibility of handle the interaction with the service.namespace HelpDeskBot.Dialogs { using System; using System.Collections.Generic; using System.Threading.Tasks; using Microsoft.Bot.Builder.Dialogs; using Microsoft.Bot.Connector; using Services; [Serializable] public class UserFeedbackRequestDialog : IDialog<object> { private readonly TextAnalyticsService textAnalyticsService = new TextAnalyticsService(); public async Task StartAsync(IDialogContext context) { } } }
  5. Replace the StartAsync method’s implementation to ask the user to provide feedback about the bot.public async Task StartAsync(IDialogContext context) { PromptDialog.Text(context, this.MessageReceivedAsync, “Can you please give me feedback about this experience?”); }
  6. Add a new method called MessageReceivedAsync. This method receives the user’s response and sends it to the Text Analytics API to evaluate the user sentiments. Depending on the response (greater or lower than 0.5) a different message is displayed to the user.public async Task MessageReceivedAsync(IDialogContext context, IAwaitable<string> result) { var response = await result; double score = await this.textAnalyticsService.Sentiment(response); if (score == double.NaN) { await context.PostAsync(“Ooops! Something went wrong while analyzing your answer. An IT representative agent will get in touch with you to follow up soon.”); } else { string cardText = string.Empty; string cardImageUrl = string.Empty; if (score < 0.5) { cardText = “I understand that you might be dissatisfied with my assistance. An IT representative will get in touch with you soon to help you.”; cardImageUrl = “”; } else { cardText = “Thanks for sharing your experience.”; cardImageUrl = “”; } var msg = context.MakeMessage(); msg.Attachments = new List<Attachment> { new HeroCard { Text = cardText, Images = new List<CardImage> { new CardImage(cardImageUrl) } }.ToAttachment() }; await context.PostAsync(msg); } context.Done<object>(null); }NOTE: For sentiment analysis, it’s recommended that you split text into sentences. This generally leads to higher precision in sentiment predictions.

Task 3: Modify the Bot to Ask for Feedback and Analyze the User’s Sentiments

  1. Open the RootDialog.cs in the Dialogs folder. Locate the IssueConfirmedMessageReceivedAsync method. Update the code block when the user confirmed the ticket to call the UserFeedbackRequestDialog dialog and ask the user for feedback. Also move the context.Done<object>(null); line inside the last else. The resulting code should look as follows.private async Task IssueConfirmedMessageReceivedAsync(IDialogContext context, IAwaitable<bool> argument) { var confirmed = await argument; if (confirmed) { … if (ticketId != -1) { … } else { await context.PostAsync(“Ooops! Something went wrong while I was saving your ticket. Please try again later.”); } context.Call(new UserFeedbackRequestDialog(), this.ResumeAndEndDialogAsync); } else { await context.PostAsync(“Ok. The ticket was not created. You can start again if you want.”); context.Done<object>(null); } }

Task 4: Test the Bot from the Emulator

  1. Run the app clicking in the Run button and open the emulator. Type the bot URL as usual (http://localhost:3979/api/messages).
  2. Type I need to reset my password and next choose a severity. Confirm the ticket submission, and check the new request for feedback.
  3. Type It was very useful and quick. You should see the following response, which means it was a positive feedback.
  4. Repeat the ticket submission and when the bot asks for feedback, type it was useless and time wasting. You should see a response as follows, which means it was a negative experience.

Leave a Comment