Visualize Your Sentiment Analysis in Tableau

Visualize Your Sentiment Analysis in Tableau

Sentiment analysis is the automated process of identifying emotions in text. quickly make sense of opinions ‒ like those in social media posts, surveys, product reviews, and support conversations ‒ and understand how customers feel about your business. 

Try out this free sentiment analyzer to see how it transforms your texts into quantifiable insights, ready to import to Tableau.

Why Use Sentiment Analysis & Tableau?

Picture this: your company has recently launched a mobile app and wants to analyze user feedback in online reviews. With sentiment analysis, you can avoid manually tagging hundreds of reviews and instantly get a spreadsheet with all these comments tagged as positive, negative, or neutral.

To present your sentiment analysis findings to co-workers and stakeholders, you’ll want to use data visualization tools like Tableau, a business intelligence software that allows you to create powerful and engaging interactive dashboards.

In just a few clicks, you can connect your sentiment analysis spreadsheet to Tableau and visualize the number of positive and negative opinions about your mobile app, as well as insights, trends, and patterns. If you want to dive deeper into your analysis, you can perform aspect-based sentiment analysis, which tells you what customers are saying about specific features: your app’s performance or usability, for example?

Ready to get started? Below, we’ll walk you through how to perform sentiment analysis with MonkeyLearn, and show you how to connect your data with Tableau so you can create engaging and easy-to-understand reports.

How to Do Sentiment Analysis?

MonkeyLearn is a user-friendly machine learning platform that allows you to perform sentiment analysis using pre-trained models or customized sentiment analysis models.

With pre-trained models, you can start doing sentiment analysis right away (no coding skills required). Just paste a text into this online sentiment analyzer and see how it sorts your text data in a matter of seconds!

Remember though, this is a generic sentiment analysis tool. If you want to tailor machine learning tools to your business for even more accurate results, you’ll need to build a custom model. 

Build a Custom Sentiment Analyzer in 6 Easy Steps

With MonkeyLearn, creating a custom sentiment classifier only involves a few steps:

1. Choose your model

Sign up to MonkeyLear for free, go to the Dashboard, and click on Create a model. Then, choose “Classifier”.

2. Choose a classifier

To create a model that can classify text by sentiment, click on “Sentiment Classification”.

3. Import your data

First, you’ll want to run your sentiment data through the opinion unit extractor, which separates text into opinions for more accurate results. For example, one review might contain both positive and negative opinions.

Opinion extractor separating opinions

Next, upload an Excel or a CSV file with your opinion units, which will be used to train your sentiment classifier.

4. Train your sentiment analysis model

Train your sentiment analysis model by manually tagging each piece of text as positive, negative, or neutral. Once you’ve tagged a few examples, the model will start assigning sentiment scores and making its own predictions. You can re-tag inaccurate examples to improve your model’s performance.

Tagging training data.

5. Test your sentiment classifier

Go to the “Run” tab and paste a text to test your model. You can always tag more examples manually to improve accuracy.

Test with your own text



6. Put your model to work!

Once you’ve finished training your sentiment classifier, you can move forward to analyzing large-scale data. There are three options for this: 

  • Batch processing: upload an Excel or CSV file with your data. In return, you’ll get a file with your sentiment analysis results. 
  • Integrations: you can connect to a third-party app like Zapier, Zendesk, or Google Sheets.
  • MonkeyLearn API. Developers can manage models programmatically through the easy-to-use API.

Visualize Sentiment Analysis Results with Tableau

So you analyzed your data and received a CSV, Excel, or JSON file in return. What next? With Tableau, you can organize your sentiment analysis results and create effective and powerful data visualizations. Just follow these steps:

1. Request a free trial and install Tableau.

Click on “Try now” to access a 14-day free trial of Tableau Desktop. Download and install the package. 

2. Connect to a data source.

To start visualizing your data using Tableau, you need to connect to a data source. Tableau supports a wide number of files and databases, like Excel, Google Sheets, CSV, JSON, and more.

If your data is stored in an Excel file, for example, go to the menu on the left (“Connect”), find the option “To a File” and select “Microsoft Excel”. 

3. Choose your file and sheet

Now that you’ve connected your Excel file with Tableau, you’ll have to choose the file that contains the data you’d like to visualize ‒ for example, a file with app store reviews tagged by sentiment and aspect ‒ and specify which sheets to use. 

Once you’ve done this, Tableau will show you a preview of the data you are selecting:

4. Start building charts and graphs

Access the workspace by clicking on “Sheet 1” in the toolbar at the bottom:

In the workspace, the left column shows two parameters that you need to set up: “Dimensions” and “Measures”

Dimensions and measures organize your data in Tableau and are the structure the program will use to create visualizations. To pick the ones you’d like to use, double click and they’ll be placed as columns or rows. 

In this example, we’ll use Aspect as “Dimension” (rows) and Positive, Negative, and Neutral as “Measures” (columns).

Then, head to the column on the right-hand side (“Show me”), which displays all the visualization options available for the dimensions and measures that you’ve just selected:

Display all the visualizations.

Select your visualization format and see the results!

The above example shows a horizontal bar chart that presents the number of positive, neutral, and negative opinions (the “measure” values) for each aspect (dimensions). 

You can change the visualization type any time you want, by choosing a different option from the “Show me” toolbar: 

5. Bring your visualizations together in a Dashboard

Tableau allows you to combine different charts and graphs into a “Dashboard”. Instead of creating different spreadsheets (like a presentation), you can have all your data displayed on a single screen, allowing you to create insightful stories. 

To create a dashboard, go to the top menu bar and click on “Dashboard” / “New Dashboard”. Then, drag the sheets you’d like to add to your dashboard. 

The example below shows a dashboard with three different ways of visualizing the results of your sentiment analysis.

If you want to keep learning how to do awesome visualizations using Tableau, check out these training courses from Tableau.

Final Words

Sentiment analysis gives you insight into the things that customers like and dislike about your brand and products. However, to fully explore the possibilities of this text analysis technique, you need data visualization tools to help you organize the results.

With tools like Tableau, you can connect to a variety of sources and easily make sense of your data. Interactive dashboards allow you to spot trends and patterns that you may have otherwise missed, and uncover precious insights you can share with your teams to improve specific areas within your business. 

Combine the power of data visualization with MonkeyLearn, a machine learning platform that makes sentiment analysis very straightforward. You can choose to use a pre-trained solution or build your own custom model following a few simple steps. Sign up to MonkeyLearn for free and start getting value from your data!

Rachel Wolff

May 5th, 2020

Posts you might like...

MonkeyLearn Logo

Text Analysis with Machine Learning

Turn tweets, emails, documents, webpages and more into actionable data. Automate business processes and save hours of manual data processing.

Try MonkeyLearn
Clearbit LogoSegment LogoPubnub LogoProtagonist Logo