Even before digital screens became ubiquitous in society, the accessibility of texts and being able to access information has been a challenge. The emergence of digital screens and new technologies has certainly changed how people of different abilities access information, such as the use of screen readers. There are also various principles of design that come into play when displaying information on a digital screen or in print, principles designed to enhance the readability of text and to make information accessible to people with a range of needs.
A new study conducted by a team from the University of Central Florida and Adobe researchers describes an artificial intelligence tool, Adobe’s AI tool FontMART, that can personalize the fonts on a screen and adapt them to a specific person’s needs, making the information more accessible to read. The study is published in ACM Designing Interactive Systems 2022.
The researchers worked with a range of different experts: people specializing in vision, typography, data, and so on, to study FontMART.
The goal of FontMART is to build on a discrepancy noted in prior research between a person’s preferred font and the font that can improve how quickly they read and their reading experience. To address this discrepancy, researchers used the Font Preference Test, which can use certain information about a reader to make recommendations about a more appropriate font.
Findings from the study suggest these predictive abilities can help increase reading speed.
FontMART was initially validated using a 250-person cohort. Combined with expert insights from topographers, eight different fonts were selected for use in the study. These were the fonts recommended to potential readers.
Using these insights, FontMART draws on a reader’s self-reported characteristics and demographic information, including their reported reading speed and age. Researchers found that age played the biggest role when it came to selecting an optimal font for a reader.
Fonts commonly selected for people who were older included “heavier” fonts, which look bolder on a screen and are generally easier to read than lighter fonts.
In the future, researchers hope to study a wider range of ages to get a more representative sample, and even study how FontMART’s recommendations may change for different kinds of on-screen content, such as long and short form.
Sources: EurekaAlert!; UCF; ACM Designing Interactive Systems 2022