DysHelper: The Dyslexia Assistive User Experience
Tereza Pařilová, Romana Remšíková · 2018 · Proceedings of the 15th International Web for All Conference (W4A 2018) · doi:10.1145/3192714.3196320
Summary
This demonstration paper presents DysHelper, a web browser extension designed to assist people with dyslexia by visually modifying the appearance of web text without altering its content or meaning. The extension addresses the core visual processing difficulty in dyslexia: problematic graphemes (letter shapes) that are similar or close to each other cause overlapping, switching, or rotation during reading. Unlike other solutions such as text-to-speech (which bypasses reading entirely and may negatively impact children’s reading development) or the DysWebxia service (which removes pseudo-unnecessary elements like commercials and formatting), DysHelper takes a unique "fragmentation" approach. It visually separates problematic letter patterns by inserting a fragmenting sign (resembling a hyphen) between graphemes that are likely to be confused, pulling them apart while keeping the text visually consistent. The extension was developed at Masaryk University in Brno, Czech Republic, motivated by the significant gap in dyslexia research for non-world languages like Czech, Polish, and Swedish — where most empirical studies focus on English and Spanish. Dyslexia is recognised as affecting up to 20% of populations using the Latin alphabet, resulting from miscommunication between the left and right hemispheres in the corpus callosum, making it highly individual because the information flow is unsystematic and asynchronous. The extension is based on individual tests that identify which specific letter combinations are problematic for each user, followed by personalised modification of web-based text.
Key findings
A small user experience study was conducted with university students over 18 who had been clinically diagnosed with dyslexia, though only 5 students completed all stages. The study consisted of individual user testing where participants read two types of texts (a factual text from Wikipedia and a narrative fairy tale), followed by discussion. Participants were asked to evaluate DysHelper on a 1-10 scale across multiple dimensions. The extension received a mean rating of 8 for "suitable for daily use" and 7.6 for "would you recommend it." When asked about impact, participants rated it 7.8 for "does it improve perception of written text" and 7.4 for readability improvement. Participants noted that the fragmentation approach helped them perceive individual letters more clearly and reduced the visual confusion caused by similar letter shapes. However, the study also revealed that individual needs varied significantly: what helped one user could be neutral or even distracting for another, reinforcing that dyslexia is highly individual and that assistive tools must be personalisable. The fairy tale text was generally easier to read with the extension than the factual Wikipedia text, possibly because narrative text has more predictable sentence structures. Participants appreciated that the extension modified text appearance without changing content, allowing them to read actual web content rather than simplified versions.
Relevance
DysHelper represents an unusual approach to dyslexia assistive technology — rather than changing fonts (like OpenDyslexic), simplifying content, or reading text aloud, it targets the specific visual confusion between similar graphemes by physically separating them. This addresses the reading process directly rather than bypassing it, which is important for maintaining and developing reading skills over time. For accessibility practitioners, the key insight is that dyslexia assistance need not mean replacing reading with listening; visual modifications that reduce specific perceptual confusions can support actual reading. The individualised approach — where the extension identifies each user’s problematic letter combinations through testing — acknowledges the heterogeneity of dyslexia, which is often overlooked by one-size-fits-all solutions. The focus on Czech language is valuable given the dominance of English in dyslexia research. As a 2-page demonstration paper with only 5 participants, this is preliminary work, but the fragmentation concept and personalisation approach merit further investigation with larger samples and controlled reading performance measures.
Tags: dyslexia · browser extension · reading accessibility · typography · web accessibility · text adaptation · Czech language · cognitive accessibility · personalization