Have you ever thought about what the thousands of reviews given by people on social media about your newly released product mean? Unfortunately, all these reviews cannot be categorized as either good or bad. One needs to truly understand what consumers feel about your product or business, and this is where sentimental analysis comes in.

In this blog, we will look at the functionality, highlights, and limitations of doing sentimental analysis across Tableau and Power BI.

Use Case for Sentimental Analysis

When looking to improvise their products/services, every organization needs to analyze customer feedback from various sources like surveys, social media platforms as well as verified reviews from app stores. Feedback collected from all the sources will be in unstructured text format. But with the perks of sentimental analysis, it has become easier to organize unstructured data with its related emotions to derive insights, which leave us with the best use case for sentimental analysis.

Example for Sentimental Analysis

In our example, we are using Customer reviews for Amazon Echo, Firestick, Echo Dot to perform our sentimental analysis. Let us now study how it is done across Tableau and Power BI.

Tableau

Tableau enables us to do sentimental analysis on our data after we integrate Tableau with Python. Once the integration is done, we can create a calculated TabPy field. TabPy is a little different from normal python code. Once we are clear with the basics of TabPy coding, we can create a calculated field to find out the sentimental score of each review. Sentimental scoring can be done by following these steps.

Figure 1-TabPy calculated field
Figure 2-Sentimental scoring

Advantages

  • With TabPy Script, any algorithm of the user’s choice can be used.
  • Any package of the user’s choice can be used for maximum accuracy.

Limitations

  • The load time for a chart that uses the Tableau calculated field increases when the dataset is very large.
  • TabPy calculated fields will not be extracted if a Tableau extract is created.
  • TabPy calculations cannot be used as a reference to create another calculated field unless both the TabPy field and additional field are used together in the view.

Power BI

With Power BI, sentimental scoring (opinion mining) or key phrase extraction can be done easily using Azure Cognitive services which use inbuilt machine learning algorithms.

Given raw unstructured data, it can extract data and analyze sentiment or extract the most important key phrases with just a click.

Real-time data can also be fetched in Power BI using the web option to get data. Python scripts can also be used to get sentiment scores wherein we can use algorithms of our choice.

Figure 3- Sentimental scoring using Azure cognitive services

Advantages

  • With Python script, any algorithm of the user’s choice can be used. (Azure Cognitive services uses inbuilt machine learning algorithms)
  • No gateway is required for data refresh in Azure Cognitive services whereas a personal gateway needs to be installed while using Python script.

Limitations

  • Python script gives a timeout error after 5 mins of execution.
  • Azure Cognitive services support only data in import mode.
  • Currently, Python scripts can’t be used in reports with enhanced metadata preview feature enabled.
  • There are certain data and rate limits while using Azure Cognitive Services.

We hope this blog helped you as a guide to implementing sentimental analysis using multiple BI Tools. To learn more about Visual BI’s Consulting & End User Training Programs, contact us here.


Corporate HQ:
5920 Windhaven Pkwy, Plano, TX 75093

+1 888-227-2794

+1 972-232-2233

+1 888-227-7192

solutions@visualbi.com


Copyright © Visual BI Solutions Inc.

Subscribe To Our Newsletter

Subscribe To Our Newsletter

Join our mailing list to receive the latest news and updates from our team.

You have Successfully Subscribed!

Share This!

Share this with your friends and colleagues!