Getting your Trinity Audio player ready...
|
Digital technology has revolutionised how the world interacts with data in recent years. From complex visualisations to interactive charts, these tools have empowered users to explore and analyse data in previously unimaginable ways. However, for people who are blind or have low vision, accessing and understanding these visual representations can be a significant challenge.
Most tools for creating data representations rely on converting visual charts into accessible formats. While this approach has enabled some level of accessibility, it also creates barriers for blind and low-vision users. These users are often limited to exploring existing visualisations and cannot make custom representations. This limitation restricts their ability to analyse and analyse data and hinders their participation in data-driven discussions and decision-making processes.
Recognising these challenges, a team of MIT and University College London (UCL) researchers set out to develop a new approach to accessible data representations. Their goal was to create a software system enabling blind and low-vision users to build custom, multimodal data representations without needing an initial visual chart. The result of their efforts is Umwelt, an innovative software system designed to revolutionise the way blind and low-vision users interact with data.
Umwelt, which means “environment” in German, is an authoring environment designed explicitly for screen-reader users. It features an intuitive editor that allows users to upload a dataset and create customised representations, such as scatterplots, using three modalities: visualisation, textual description, and sonification. Sonification, a process that converts data into nonspeech audio, enables users to perceive data patterns through sound.
One of the critical features of Umwelt is its ability to represent various data types, ranging from simple numerical data to complex multivariate datasets. This flexibility allows users to explore and analyse data in ways that are tailored to their individual needs and preferences. For example, a user could create a scatterplot that represents stock prices using different tones to represent the price for each date, arranged by ticker symbol.
To test the effectiveness of Umwelt, the researchers conducted a study with five expert screen-reader users. The participants found Umwelt valuable and easy to learn, praising its intuitive interface and powerful features. They noted that Umwelt could facilitate communication between people who rely on different senses, enabling more inclusive and collaborative data analysis.
Jonathan Zong, the Author of the paper introducing Umwelt, emphasised the importance of enabling blind and low-vision individuals to participate fully in data analysis. He highlighted Umwelt’s potential to shift researchers’ perspectives on accessible data analysis, emphasising the importance of multisensory approaches.
“This technology has the potential to not only improve the accessibility of data analysis tools but also to enhance the overall quality of data analysis by incorporating diverse perspectives and modes of understanding,” Zong said.
In addition to its practical applications, Umwelt has the potential to drive scientific inquiry into how people use and perceive multimodal representations. The researchers plan to release an open-source version of Umwelt to enable further development and integration of new modalities, such as tactile sensing, into the software system.
“This new tool not only opens up possibilities for individuals with visual impairments to engage more fully in data analysis but also challenges us to rethink the fundamental assumptions underlying data representation and accessibility,” said Arvind Satyanarayan, Senior Author of the paper and Associate Professor of Computer Science at MIT.
Umwelt represents a significant advancement in making data analysis more inclusive and accessible to individuals with visual impairments. He believed that by empowering blind and low-vision users to create their custom data representations, Umwelt could revolutionise how people think about accessible data analysis and open up new opportunities for collaboration and innovation in data science.