AI Model Recommends Personalized Fonts to Improve Accessibility

October 10, 2022

Research from the University of Central Florida (UCF) suggests that personalized fonts might be able to improve digital accessibility by enhancing reading speed without reducing reading comprehension. 

The study was a collaboration between UCF and Adobe. Participants completed a Font Preference Test, which enabled a machine learning model to create a unique typography for each reader. 

“The future of readability is a device watching humans read and using their performance to tailor the format so that they read at their best,” said Ben Sawyer, director of the Readability Consortium and UCF’s Virtual Readability Lab. “We look forward to the day when you can pick up a device, read and receive information in a way that uniquely suits your needs.”

The technology could have profound implications for the future of website accessibility. However, as we’ve discussed in other articles, the link between font choice and reading comprehension is complex — and more research is needed to determine whether UCF’s model could meaningfully improve reading speed for all readers. 

Font preference is a hot topic in the digital accessibility space 

UCF’s research isn’t the first attempt to address accessibility concerns through typography. 

Dyslexia fonts, which attempt to improve experiences for people with dyslexia by using distinct characters, have become a popular choice in recent years — but there’s not much scientific evidence to support the idea that typeface affects reading speed or comprehension. 

However, UCF’s study goes much further than traditional “accessibility fonts" by using artificial intelligence (AI) to choose specific font characteristics for different individuals. 

“FontMART can predict the fonts that work well for specific readers by understanding the relationship between font characteristics and reader characteristics like font familiarity, self-reported reading speed, and age,” UCF’s Cara Cannon wrote

"For instance, font characteristics like heavier weight benefit the reading experience of older adults because thicker font strokes are easier to read for those with weaker and variable eyesight.”

Related: The U.S. Population Is Aging. Is Your Website Accessible for Older Adults?

How UCF Researchers Created Personalized Font Recommendations Through Machine Learning

UCF used Adobe’s FontMART, a learning-to-rank model that automatically orders a set of eight fonts by predicted reading speed. The options, all of which are freely distributed or available on both Windows and MacOS operating systems, included four serif fonts and four sans serif fonts: 

  • Times (Serif)
  • Source Serif Pro (Serif)
  • Georgia (Serif)
  • Merriweather (Serif)
  • Roboto (Sans Serif)
  • Aria (Sans Serif)
  • Open Sans (Sans Serif)
  • Poppins (Sans Serif)

According to the study’s abstract (PDF), FontMART’s fastest font prediction showed an average increase of 14-25 words per minute (WPM) compared to other font defaults. Faster reading speeds did not reduce reading comprehension.

Participants were also asked to identify their font preferences. Interestingly, only about 30% of participants read fastest in their preferred fonts. In other words, while people preferred the appearance of certain fonts — in particular, Merriweather — they were able to read significantly faster with FontMART’s suggestions.

Related: 5 Quick Ways to Self-check the Accessibility of a Website

A Quick Guide to Accessible Fonts

The Web Content Accessibility Guidelines (WCAG), widely considered the international standards for digital accessibility, do not recommend a specific font for online content. That’s partially because font choice — while important for aesthetic reasons — seems to be much less important for accessibility than other aspects of typography.

For web content, most accessibility experts recommend popular fonts with a clean, sans serif aesthetic. Serif fonts have “ticks" and “tails" that take up a larger amount of space on a screen than they do on a printed page, and sans serif fonts generally display better on different types of computers and mobile devices.

In our article, Best Fonts To Use for Website Accessibility, we recommend a few simple guidelines for improving readability: 

  • Choose an appropriate font size. Most websites display normal text at a size of 12 to 14 points (pts). This can be increased to improve legibility.
  • Avoid using images of text, which can impact people who use screen readers and other assistive technology (AT). 
  • Try to use as few fonts as possible. Too many fonts can reduce legibility and cause confusion.
  • Try to avoid using the appearance of a font (such as bold or italic formatting) to convey meaning. Some screen readers may not inform users when text is bolded or italicized. 
  • Make sure your font color maintains appropriate color contrast with its background. The Bureau of Internet Accessibility offers a free Color Contrast Accessibility Validator for checking color-pairs against WCAG recommendations.

For more guidance, send us a message or check your website against WCAG with a free automated analysis.

Receive an industry accessibility analysis of your website

Recent posts

6 Free Tools for Evaluating Web Accessibility

January 25, 2023

What to Know Before Writing a WCAG Conformance Claim

January 24, 2023

What Are Cognitive Disabilities, and Why Are They Important for Web Accessibility?

January 23, 2023

Not sure where to start?

Start with a free analysis of your website's accessibility.

GET STARTED