Text Analytics Playground Features

1. Generate topics and insights for any unstructured text dataset
2. Leverage pre-trained ML models to generate topics and insights for unstructured text datasets
3. Define topics and keywords for your unstructured text dataset
4. Visualize keyword and topic distribution and overall contribution
5. Export visuals with data insights to a PDF format

1. Open the Text Analytics Playground by clicking here
2. Sign yourself into the portal by providing the necessary consent and clicking “I Agree”


Figure 1: Login Page

3. Once consent is provided, you will be redirected to the application’s Data Connect page


Figure 2: Data Connect landing page

4. On Data Connect, you can provide text input using any of the following options:
   • Click “Use Sample Dataset” (in-built)
   • Upload a file (the playground supports xlsx and csv)
   • Type the comments directly into the text box (use the Enter key as a delimiter)


Note: The Dataset Category dropdown (bottom-left, below the textbox) determines the visualizations shown in topic modeling. If the dataset comes from any event related data, the visualizations will be different from a dataset that falls under “Other Data”. Select the option most relevant to your data.

5. Once you finish providing all the necessary information, you can generate insights by clicking “Generate Insights”. You will be taken to the Insights page
6. On the Insights page, you will have access to the following visualizations:
   • Number of Responses: the number of line items provided in the file or entered as text
   • Number of Topics: the number of topics associated with the data or the number of topics selected by the user
   • Average Contribution Score: the average of all the contribution scores of the topics
   • Topic Strength: the percentage importance of the topic in the data
   • Topics: all the topics involved
   • Common Categories: the number of comments tagged per topic
   • Bigram-Trigram Topics under Categories: the different two or three-word topics, along with the probability score of their being used
   • Word Cloud: the most used words associated with the topic(s)
   • Strongest Topics: based on the average contribution score of the topic
   • Word Count and Importance of Topic Keywords: includes keyword weightage
   • Topic Distribution over Response: topic distribution over the uploaded text data
   • Feedback Comments: the most prominent topic per comment along with its probability score
   • All Keywords under Category: all the keywords under the selected topics
   If you chose the “Other Data” option, you will be provided with additional visualizations:
   • Topic Keyword Distribution Chart: the keyword distribution (this lets you analyze them in order to interpret the topics)
   • Best Response to Describe Topic: the text that has the highest correlation with the selected topic
7. The tool also gives you the option to modify topics. Go to the “Looking For Something Else” section. There are two ways to create a new topic:
   1. Create Topic:
      You can add topics of your choice by providing the keywords and the topic associated
   

   Figure 3: Create Topic allows you to create custom topic-keyword mapping

   2. Filter Topic:
      You can choose or eliminate the topics shown to suit your requirements
   

   Figure 4: Filter Topic provides you the topic filtering capability

8. Once you finish, re-run the model. You will be able to see the changes in the visualizations
9. To download the insights generated into a shareable PDF, click the Download button

Acquiring the Tool

To discuss customization and prices, please contact our sales team at Sales@MAQSoftware.com.

Feedback

To get help on a specific question or problem please contact us at Support@MAQSoftware.com.