Despite the ease in Covid-19 restrictions over the past year, people (including myself) are still spending more time than ever online. The increased usage and reliance during the pandemic has now made the world wide web (WWW) an even more indispensable part of our lives. And with new innovations like Web 3.0 moving into the mainstream, this is not going to change anytime soon. On the face of it, however, when we look at the design and visualisation of the information resources and services on the WWW today, it seems like nothing has moved on. Even though the functionality, with its endless data and new interactions, has greatly advanced from web 1.0 to now web 3.0. The user interface design of the web still seems to be in a time warp (i.e. still mainly concerned with the usability issues identified in the late 1980s). Specifically, the emerging issues of today around online safeguarding and inequality are still not fully being addressed. In this keynote talk I will briefly discuss the ‘usability’ evolution of the WWW. The focus will be on the current design of applications on the WWW and how their user experience design (UXD) needs to evolve. I will use examples from different projects to demonstrate how we can use new visual approaches to design for more effective UXD and the greater wellbeing of all users.

Visualisations are used in various context from scientific articles to data journalism for communicating data and/or algorithmic results, and as evidences to support and contextualise arguments. While the design and the visual elements in a visualisation determine what a visualisation encodes, a visualisation operates once viewed and interpreted by its viewers. Designing visualisations requires therefore a better understanding and consideration of such perceptual, cognitive and interpretive processes. To that end, there is growing interest within the visualisation research community in empirically studying how visualisations are perceived, and whether design decisions could be informed by understandings emerging from these studies. This talk will introduce this relatively young research area, and discuss our own research involving studies of how statistical concepts such as spatial auto-correlation, correlation and more complex computational models are understood through visualisations. We will discuss the potential for methodological innovation, for instance, the use of natural language understanding in gaining deeper understandings of how visualisations of statistical constructs are interpreted, as well as opening up a debate around the limitations of how we currently run such empirical studies.