Visual Analytics: The need, model, tools and applicationsSteve, VP of an MNC looking at massive spreadsheets containing sales, revenue, spends, cost heads by regions, product etc wondering, “How great it will be if I get a crisp and concise view of all this data every morning. It will work wonders in optimizing my business and boost the bottom line. And the icing on the cake would be if I could zoom into any region of interest to check what sells and what not…” What Steve didn’t know he needed was Visual Analytics (VA).

What is Visual Analytics?

Just like a picture speaks thousands of words, Visual Analytics can talk thousands of data points. Statistical methods are difficult for an ordinary person or business users to understand and correlate with a business problem in order to make a decision and hence the need is for visualization. In layman’s language, visual analytics is to visually represent the information, with the capability to allow human direct interaction with the tool to produce insights, to draw conclusions, to slice–dice information in real time and to facilitate better decisions.

The Need for Visual Analytics (VA)

The mother of invention is the need. This new age digital era is full of data and which is exponentially increasing with websites, smartphones and smart technologies connected to the internet. Data is captured like never before. The amount of data generated every minute is mind-boggling. With ever-increasing competition and shrinking margins, using data to take the better decision is serving a differentiator. This has created a need to look for visual-based data analysis that can make sense out of the huge data.











In the field of Business intelligence or data analytics, the insights in visual form are tremendously helpful. It makes the consumption of insights easy to understand and take decisive action. This creates a need for data visualization tools to process and interpret the complex data sets and models.

How can a Visual Analytics Model help?

Visualization helps to perceive new aspects of the data. The user can explore the data models and achieve new knowledge by incorporating interactivity in visualization.

Basic to any analytics is to get the right set of data. It starts with data cleaning, normalization, grouping, or integration of heterogeneous data sources. Then the analyst needs to evaluate the appropriate modelling technique to cater to the problem. Model visualization can then be used to create evaluate the findings of the generated model. It is a herculean task working with hundreds of variables for creating a model like logistic.  Visual analytics approach to modelling for high-dimensional data sets can simplify things. It leverages traditional modelling by providing intuitive visualizations for inspecting statistical indicators. Finally, the knowledge and insights gained from visualization are used to take analytical or outcome driven business decisions.

Further growing need for Big Data

Visual Analytics is especially indispensable while dealing with a large chunk of complex data, these days typically referred as Big data. Big data is data sets that are so voluminous and complex that traditional data processing application software is inadequate to deal with them. The properties of Big Data that makes it difficult to handle using conventional techniques-

Volume: Massive Volume of Data

Velocity: Speed of data to be stored, retrieved and analyzed

Value: Ability to turn our data into value

Variety: Data comes from all streams structured, unstructured and semi-structured

Veracity: The data being stored and mined are meaningful to the problem being analyzed ie trustworthiness of data.

Visual Analytics in Criminal Investigation

One of a very useful application of visual analytics is in identifying criminal linkages by the investigation agencies. For example, by looking at the movement of funds between suspects involved in activities like drug trafficking, bomb blasts etc the actual Kingpin or the mastermind can be identified. This might not be a difficult task by looking at the raw data or summarized statistics.


Visual Analytics in the Retail industry

Retailers are leveraging visual analytics to understand and make sense from large piles of invoices/bills generated at each of their stores. It helps them answer questions like-

Plan for inventories- L size shirts sells them most, why don’t stock them more and less of others?
Decide Product Mix- Should we keep dairy products or only other FMCG products
Stock by season- Should we stock sweaters in summer? In winter should we have same stock levels same in Mumbai and Delhi?
Market- Should we stock rice in same volumes in south and north region?
Compare stores- Which region store is more profitable?

It also helps to look at the larger picture of operations for a sustainable and profitable growth.

Other Applications:

Some of the other applications of visual analysis are:

  • Financial Analysis
  • Network Security
  • High-dimension & Subspace Analysis
  • Document Analysis
  • Molecular Biology

Tools in use for Visual Analytics

Tableau is considered one of the best data visualization platforms, it’s simple, drag and drop based interface gets it the major brownie point. The basic version of Tableau Desktop comes free for students and instructors at accredited academic institutions, whereas users have to pay for advanced features.

One of the most traditional vendors in advanced analytics space, SAS visual analytics has been in the industry for a long time offering insights to various businesses. It’s capable of delivering fast answers to complex questions regardless of the size of the data.

R can also have packages like shiny for interactive dashboards. Shiny is designed for fully interactive visualization, using JavaScript libraries like d3, Leaflet, and Google Charts. It allows a user to create an interface that changes dynamically.

For NSE certified course in Business Analytics click here & for PG Diploma in Data Science click here: