Information Visualisation

Information visualisation transforms abstract data into graphical representations that leverage human visual perception to support analytical reasoning. By mapping data attributes to visual variables—such as position, colour, size and shape—researchers reveal patterns, trends and anomalies that remain hidden in tabular or textual form. Modern visualisation embraces both static graphics and interactive environments, guiding users through large-scale and high-dimensional datasets via coordinated views, dynamic filtering and details-on-demand. Foundational principles such as clarity of encoding, minimisation of visual clutter and perceptually uniform colour scales ensure that displayed information remains both accurate and accessible. From simple bar charts and scatterplots to complex network diagrams and spatial projections, the field has expanded to include specialised techniques for temporal data, multivariate embeddings, graph structures and geospatial maps. Today’s tools integrate automated layout optimisation, web-based platforms and machine-assisted suggestions, enabling domain experts and general audiences alike to explore data in domains ranging from genomics and environmental science to finance, public policy and media. By balancing rigorous data processing with user-centred design, information visualisation continues to play a vital role in insight generation, decision support and the democratisation of knowledge.

Research from Nature Portfolio

One recent contribution revisits popular non-linear dimensionality-reduction techniques, demonstrating how systematic exploration of parameter settings and diagnostic plots can mitigate misleading artefacts in methods such as t-SNE and UMAP, thereby improving the interpretability of high-dimensional projections. A separate study introduces a spectral assessment framework that assigns an eigenscore to each data point’s representation across multiple visualisations and then fuses these views into a consensus display, enhancing the preservation of local and global structure by combining complementary algorithmic strengths. Another work examines common colour-scale practices in scientific communication, highlighting how uneven gradients and red–green mappings can distort quantitative variation or exclude viewers with colour-vision deficiencies; it provides guidelines for adopting perceptually uniform palettes that faithfully convey data and maximise accessibility.

Research from all publishers

A new web application streamlines the creation of circular genomic and network diagrams by integrating automated layout optimisation, modular templates and real-time parameter controls, allowing non-programmers to generate publication-quality Circos plots through an intuitive graphical interface. Advances in colormap design have produced perceptually uniform, colour-vision-deficiency-safe palettes by optimising colour ramps within modern appearance-based colour spaces, ensuring that both typical and CVD viewers can interpret scientific data with minimal perceptual bias. Furthermore, a comprehensive survey of visual analytics for machine learning has established a taxonomy of interactive techniques across the model-development lifecycle—spanning data preparation, model building and post-hoc evaluation—and showcased representative systems for feature importance analysis, hyperparameter exploration and fairness diagnostics, while identifying scalability and usability challenges in large-scale applications.

Information Visualisation publication trend

The graph below shows the total number of articles in information visualisation across all publications each year (not limited to Nature Index journals).

Technical terms

Non-linear dimensionality reduction: Methods that embed high-dimensional data into lower dimensions by preserving neighbourhood structure, commonly implemented by t-SNE or UMAP.

Consensus visualisation: A unified display created by quantitatively combining multiple individual visualisations to synthesise their complementary strengths.

Colormap: A sequence of colours mapped to numerical data values in a graphic, designed to convey quantitative differences uniformly and support perceptual clarity.

Interactive visual analytics: A discipline combining automated data analysis algorithms with user-driven visual interfaces to support exploratory reasoning and insight generation.

Circos plot: A circular diagram that arranges data categories or genomic elements around a ring and connects related elements via ribbons or chords to reveal complex relationships.

References

1. Seeing data as t-SNE and UMAP do. Nature Methods (2024).
2. A spectral method for assessing and combining multiple data visualizations. Nature Communications (2023).
3. The misuse of colour in science communication. Nature Communications (2020).
4. shinyCircos‐V2.0: Leveraging the creation of Circos plot with enhanced usability and advanced features. iMeta (2023).
5. Optimizing colormaps with consideration for color vision deficiency to enable accurate interpretation of scientific data. PLOS ONE (2018).
6. A survey of visual analytics techniques for machine learning. Computational Visual Media (2020).