How to Use Sentiment Analysis


In Qualaroo, sentiment analysis can quickly collect qualitative data from your text-based responses. Sentiment Analysis is powered by IBM Watson’s Natural Language Understanding to unlock critical user insights based on the following attributes:




It is a scaleable way to swiftly & accurately categorize and assign values to your qualitative data. Your text-based responses are classified into a score of 

  • 0.1 to 1: Positive

  • -0.1 to -1: Negative 

  • 0: Neutral




Analyze your qualitative data through a quantitative rating of five emotions: anger, disgust, sadness, fear, and joy. Each grouping of freeform content is assigned a value from 0 (low expression) to 1 (high expression) for each emotion.


Relevant Keywords


Isolate relevant keywords your customers use to understand which areas of your product are exceeding, meeting, or failing to meet expectations.


NOTE: Sentiment Analysis is an add-on feature supported for responses in the English language only.


Benefits of Sentiment Analysis:


  • Focus on relevant responses using automated feedback mining.

  • Import all your responses to any visual analytics platform for an enhanced understanding of user sentiments to increase conversions.


Here's what the sentiment analysis will look like:


Sentiment Analysis


In this article, you will learn:


1. How to Set up Sentiment Analysis

2. How to Access Watson Data Within the Reporting Dashboard

3. How to Use Sentiment Analysis


 How to Set up Sentiment Analysis


In the DESIGN section, 

  • Choose “Text-based answer” in the Answer type.

  • Select the checkbox in front of the “Enable sentiment analysis” option.


Text-based Answer


 How to Access Watson Data Within the Reporting Dashboard


Watson Sentiment Analysis data is accessible through the Beta Reporting Dashboard. 


Log in to see the overall sentiment, emotion ratings, and keywords in the form of a word cloud for your responses.


Sentiment Analysis


On the reporting dashboard, you can access Watson data via


1. CSV export option


In the REPORT section, 

  • Select the Date range.

  • Click Export report.


Select Date Range and Export


2. Reporting API 


Sentiment Analysis data is only available on version 1.5 of the Reporting API and Webhooks. To access the Watson data through the Reporting API, use the following URL:{SURVEY ID}/responses


Example of JSON output for both Reporting API and Webhooks


 // ...
 "answered_questions": {
   "1234567890": {
     "question_id": 1234567890,
     "question_title": "What is the reason for your answer?",
     "question_type": "text",
     "answer": "Easy to implement, super useful, doesn't obstruct user experience in any significant way.",
     "watson": {
       "sentiment": {
         "Positive": 0.4
       "emotions": {
         "Sadness": 0.1,
         "Joy": 0.1,
         "Fear": 0.1,
         "Disgust": 0.1
       "keywords": {
         "significant way": 0.1,
         "user experience": 0.1
 // ...


 Where to use Sentiment Analysis?


There are endless ways you can use Sentiment Analysis to learn and act upon your data. Here are a few use cases to spark your creativity.


Sentiment Use Cases


  •  Prompt your customers to provide open-text feedback on a product or service you offer or their experience with your organization. Use Qualaroo’s IBM Watson integration to efficiently categorize this qualitative data into positive or negative sentiment to assist your outreach initiatives. Organize your communication based on sentiment, and prioritize your efforts by focusing first on your respondents who provided negative sentiment.

  • Analyze the feedback corresponding to NPS ratings to predict future opinions and behavior. Determine which promoters provided open-text responses with negative sentiment, and reach out accordingly to maintain their status as a promoter. 

  • Conversely, use Sentiment Analysis to detect Detractors who might have provided a neutral statement and offer them extra attention to increase their opinion of your organization.


Emotion Analysis Use Case


Understand which pieces of feedback convey emotion most strongly. Consider a company that wants to increase the number of positive ratings they receive on a public forum.


By analyzing the sentiment of the feedback they receive on-site, that company can identify users whose feedback contains the most joy surrounding their offering and, therefore, more likely to promote their company on a public forum.


They can incentivize those users to leave public feedback and increase their social presence.


Keywords Use Case


Use Qualaroo’s IBM Watson integration to understand which parts of your offering are successful or need attention. Consider a company that is gathering feedback to generate a Net Promoter Score. With Key Word analysis, determine which topics are associated with Promoters, Passives, and Detractors to understand where that company is doing well and where it needs to focus its attention.


We recommend using tools like Zapier or Webhooks to process your sentiment results in real-time. 


That is all about using Sentiment Analysis.



© 2005 - 2024 ProProfs