For several years, data visualization (data visualisation, or dataviz) has enabled the media, finance, healthcare, sport and other fields to present a viewpoint on large, heterogeneous datasets through graphical representations. To communicate effectively with a specific audience, a "good" visualization requires expertise in mathematics, statistics, computer science, social science and marketing.
Since the 2000s and the rise of the Internet, data has been all around us, with our mobile phones and connected devices (televisions, cars, watches...). Most often, we simply see this data presented on a dashboard, in a mobile app, via a figure indicating a trend on a television news broadcast, or on a map during a weather forecast. Producing these displays requires extensive preliminary processing! These operations are generally carried out in cloud infrastructures capable of processing data in real time. Unless the data is first collected and transformed, it is difficult to provide users with an intelligible visualization of it.
Let your data tell a story!
Companies have therefore emerged to offer straightforward data-visualization solutions. Most commonly, corporate projects in this field are launched to create a solution tailored to managers who need to make decisions. A Qlik-BARC-Cognizant study conducted between September and October 2016 among 250 business executives in Europe found that 20% of companies use data storytelling to gain a competitive advantage. Several programs at French universities and engineering schools have been created or updated to meet this new demand.
From the user's perspective, the challenges of data visualization combined with Big Data remain the same as before: presenting information in the format best suited to the intended audience. Yet large datasets add another layer of complexity, since the "few" most important pieces of information must be made to stand out from the mass. They can be identified only through iterative workshops that place end users—the decision-makers—at the heart of the process: information needed only when a specific problem arises should not be displayed permanently, as it detracts from the overall visualization.