Evolution in data visualization
Conveying a story using data visualization has always been a huge challenge. Many visualization tools exist and the methods have become standardized. Example, bar charts, pie charts have existed long before they have been implemented IT tools. More rich visualizations such as scatter graphs, box plots have been used for more statistical analysis.
Data visualization — traditional context
The motivation for data visualization was the need to extract information from data and present it in a more consumable way. We can try to understand it better in the following dimensions:
- Where were these visualization techniques employed? In most cases, visualizations were used in business context for analyzing financial performance or in scientific contexts such as academic papers in biology, information theory etc.
- Who were the target of these visualizations? The visualizations were targeted at well trained business executives than common folks who have no training at all.
- What was the visualization conveying? Visualizations were meant more to create a report which was meant to be a part of an overall story including the causes and call to action (recommendations).
- How were the visualizations distributed? They were usually distributed as printed or online static documents. Dynamic reports could be built in many tools, but they needed training to use and were more complex to understand.
What has changed in the recent years?
The context and use has changed tremendously in the recent years:
- Where are the visualizations employed? They are being used in all industries that affect the common population such as sporting, disease control and mass transport.
- Who is consuming the data? As data is being produced and consumed in our daily life, common person also is now a big consumer of the data.
- What is the visualization conveying? The object of visualization is no longer just a report. It needs to convey a story to the consumer. The time to convey this story is also very short as it needs to grab user’s attention very quickly.
- Where is this information being conveyed? In additions to traditional reports, information needs to be conveyed using channels such as YouTube, Meta and other social media platforms.
In addition to the above changes, an important factor is the amount of information that needs to be conveyed. For example, in a sporting match analysis, the number of dimensions of a player that are captured and shown have increased tremendously. Impact of these multiple dimensions needs to be understood by the common person.
Limitations of traditional visualizations
Traditional visualization techniques has served the needs of business and academic users well for good period of time. However, there are two major problems with these methods:
Difficulty in comprehending the information — It usually takes a seasoned analyst to fully understand the complex information presented in complex visualizations. Example, someone needs to understand what percentiles are, the concept of whiskers, outliers to fully understand a box plot. Simpler visualizations such as bar charts need a lower level of training, but the information conveyed very basic.
Multi-variate datasets — Data sets usually vary by more than one or two dimensions. Example, sales numbers by product, state and time. With the advance in means of gathering information, a lot of additional dimensions are being captured and need to be analyzed. Example: the number of dimensions captured for customer demographics is increasing day-by-day based on the customer’s digital footprint (Instagram posts, Twitter posts, Likes on Facebook etc)
These can be represented into a stacked bar graph, but as the number of dimensions grow, it is harder even for a seasoned analyst to fully visualize data. Attempts have been made at comprehending the complex interplay across dimensions using dynamic reports.
Tools available to cope with the new challenges
Fortunately, advancements in technology have provided some tools to deal with the new challenges.
Technologies with lower entry barrier: Technologies such as SVG can be used to create rich content very easily. They can be developed using simple declarative languages and do not need much programming skills. These also do not need heavy and expensive IDE’s to develop and debug. Visualization tool developers can easily auto generate SVG code.
Standardization of content delivery: Unlike earlier reporting tools, most of the report delivery is now standardized around the browser. Hence, reporting tools and technologies can develop to a common platform than to specific proprietary technologies which get outdated very fast. Example: a lot of rich visualization content developed in Adobe flash is now wasted as the player is obsolete and is no longer supported.
More powerful content delivery devices: Due to increased demand from the online gaming apps, the display devices are much more powerful. These can be used to display not just two dimensional graphs but can display high quality videos and rich animations without significant display lag.
A good technique that uses the above advancements can be seen in the development of animated graphs. Animation can be used for depicting interplay of multiple dimensions in a very intuitive way. An example of such an animation is the bar chart race. It shows a three dimensional information: an item and a measure’s variance over time. Sample below:
Tool and framework support
As of this writing (in late 2022), very limited support exists in the commercial products and open source frameworks for rich animations. However, with the support for video in HTML5, lets visualization tool developers to create video reports that can be embedded into HTML pages.
Another technology that can be leveraged is SVG. It can be used to make common display devices such as browsers have become a very powerful medium to deliver rich and captivating graphics.
Frameworks such as D3.js have very good support for creating animated graphs using SVG. Few examples from D3.js:
Conclusion
Visualization requirements and applications have evolved tremendously in the past few years. Though the base technologies exist and have matured, the product and framework ecosystem is just catching up. Vendors who can cater for these requirements will have a distinct advantage over others.
Images courtesy:
References:
