Introduction

In the modern age, data analysis has crept up to the top of the Digital Transformation paradigm and Data Analyst is becoming one of the most important roles in the industry as they are involved in key business decisions and strategies because of their insights and reporting capabilities.

But what is Data Analysis? In simple words, Data Analysis is a process of extracting useful information from data using different Statistical and Mathematical methods. But Data Analysis alone cannot help. It is because after performing Data Analysis on some data, we usually store the information in CSV file, some databases, or in some other format and it is very difficult to read data from these formats, especially for non-technical people. Also, it is very unintuitive to read data from those files. To solve this problem, we perform Data Visualization. Data Visualization is the presentation of data in pictorial and graphical format. These pictorial and graphical formats are very easy to understand for even non-technical people.

To prove this point, let’s assume an excel sheet having thousands of rows of sales data. Now let’s say we just look at the data for 30 seconds, then someone asks a question like which product is giving maximum profit or which product is giving minimum profit, it will be very difficult to provide him with a satisfactory response as it is hard to get this type of information from those data just by looking at it.

But now assume, we look at a pie chart showing that same data, for 30 seconds, and then someone asks those same questions. Can we answer the question now? Yes, now it is simple to answer those question because of the simplicity of the pie chart or any other types of charts.

This is why we use Data Visualization. It a powerful technique to display data in a simpler and easier way. So, in simple worlds, we use Data Visualization to present the extracted data in a simple and intuitive way so that anyone can understand the main information that we want to tell.

Problem Actualization

Our client from Singapore, wanted a system to analyze posts from various online and offline platforms. In that project, after analyzing all posts, we needed to visualize the information we extracted by analyzing the posts. So, we needed to build several dashboards reporting and visualization. Also, the client needed several filters in the visualizations, so that they can visualize the data with different data point combinations. In other words, they wanted the visualizations to be interactive and dynamic.

Another challenge was that the visualizations should always be the latest and updated one even though the data was constantly flowing in from multiple sources.

Moreover, the client also needed the ability to export visualizations into images or pdf documents as per the use case.

Solution

We decided to use Tableau for visualization as this is one of the most popular and powerful tools in Data Visualization and is used by millions of people worldwide. Also, Tableau provides a very powerful interface for creating visualizations that looked awesome in action besides providing all the possible types of charts, that are needed for this project. Specifically, for this project another advantage of Tableau over other similar solutions is that look-wise it was far superior.

Additionally, Tableau also provides different types of filters, starting from date filters to other specific type filters that we need for this particular problem statement. We can use those filters or add new filters to make interactive and dynamic visualizations.

As data was constantly coming in, we used Tableau Server and AWS Athena to make a system which could update visualizations automatically. So now whenever our system collected some data, Tableau was rendering the visualizations automatically.
In this project, scaling was a big topic as millions of posts were collected each hour. The biggest advantage of Tableau is its insane capability to scale and handle a large amount of data, which made the job rather comfortable, so to speak.

Another comforting factor was that, for exporting the visualizations into pdf or image, Tableau by default provides the feature to export visualizations into pdf, images etc., negating the need to code anything for that.

We used Tableau Desktop specifically for creating the visualizations. The first step was to connect Tableau Desktop with AWS S3, which was done with the help of AWS Athena. Secondly, we prepared several dashboards for displaying the analyzed data. The final step was to embed all the visualizations into a web platform, for which iframes was used.

Share this post on: