The transition from two-dimensional charts to immersive three-dimensional analytics is often described as a technological development. Recent research suggests that it should instead be understood as a change in the way people interact with complex information.

The paper “Show Me Your Best Side: Characteristics of User-Preferred Perspectives for 3D Graph Drawings” offers important evidence for this interpretation. The authors studied 23 participants using a virtual-reality application to explore 36 three-dimensional graph drawings. Their objective was not simply to determine whether people liked 3D visualisations, but to understand which perspectives people considered useful and why.

The results are particularly relevant for Data2Space because they demonstrate that human judgement can be used to inform the optimisation of immersive visualisations.

The researchers found that a relatively small set of measures could explain a substantial part of the distinction between preferred and unfavoured viewpoints. Among the most consistently relevant factors were Stress, the Gabriel Ratio, Crossings, Edge-Node Overlap and Isometric Viewpoint Deviation. Combinations of these measures provided stronger separation between preferred and unfavoured perspectives than many individual measures.

This opens an important possibility for Data2Space: the system does not need to rely exclusively on the user manually finding the best perspective. Analytical models can potentially assist in identifying views that are more likely to expose important structures.

The researchers themselves point toward this possibility. They explain that combined aesthetic measures could support the automated selection of optimal viewpoints. Their findings therefore connect empirical research on human perception with algorithmic viewpoint optimisation.

For financial analysis, this could become highly valuable.

Consider a large corporate network containing hundreds of entities and thousands of relationships. Showing the entire network in VR is technically possible, but technical possibility alone does not guarantee analytical value. If the initial viewpoint hides the relevant ownership chain, users may draw the wrong conclusions or overlook important relationships.

The research explicitly highlights this danger. In interactive analysis, the choice of what to explore next can be influenced by an earlier viewpoint, particularly when that viewpoint fails to provide sufficient information for subsequent decisions. The authors therefore see value in optimised initial viewpoints.

This insight has direct implications for Data2Space. An immersive financial analysis environment can begin not with an arbitrary view of a network, but with a perspective selected to make the principal structure immediately intelligible. The user can then move through the data, changing the perspective as the analytical question changes.

The research also suggests that 3D-specific measures are essential. Traditional 2D measures remain useful, but they do not capture all the perceptual problems created by depth. Occlusion becomes more complex, and the direction of overlap can affect interpretation. The study found, for example, that edge-node overlap could be particularly misleading because it can suggest a relationship where none exists.

This is precisely where immersive financial visualisation must go beyond simply importing conventional charts into a VR environment.

Data2Space can use the spatial dimension as an analytical variable in its own right. Depth, viewpoint, separation and the visibility of structural axes become part of the analytical process.

The study also contains a striking observation about user behaviour. Participants with different levels of experience did not show substantial differences in their preferences. The authors therefore see potential for general principles that can guide viewpoint selection across users, while also acknowledging that the relationship between preference and actual task performance requires further research.

This distinction is important. Data2Space should not claim that immersive visualisation automatically improves every analytical task. The scientific evidence is more nuanced. The research demonstrates that people can consistently identify preferred perspectives and that these preferences can be related to measurable characteristics of the visualisation. It also makes clear that further work is needed to establish how these preferences translate into analytical performance.

That nuance actually strengthens the case for Data2Space.

The future of financial analytics will not be defined simply by more data, more dashboards or more sophisticated algorithms. It will increasingly depend on the ability to understand relationships within complex systems.

Data2Space represents a step toward that future by treating financial information as a structure that can be explored spatially. The research from GD 2025 provides an important scientific indication that such exploration can be made more systematic: human preferences can be measured, visual quality can be evaluated, and combinations of metrics can help identify perspectives that reveal structure.

The result is a new vision of financial analysis: not merely reading data from a screen, but entering the structure of the data and examining it from the perspectives that make its meaning visible.