Home // ACHI 2026, The Nineteenth International Conference on Advances in Computer-Human Interactions // View article

Conversational Web Browsing: Voice-Only Navigation

Authors:
Daniele Farriciello
JingHua Ye

Keywords: web browsing, voice-based, personalized, conversational, hands-free.

Abstract:
Web browsing shapes how we work and connect, yet it still relies heavily on the mouse and keyboard. For people with physical disabilities or professionals in hands-busy environments, this dependence makes accessing the web frustrating or impossible. This study presents a solution, a VoiceNav system that allows users to browse websites entirely using voice commands. The goal is to make this system work automatically on any website, bridging the gap by enabling users to browse and control websites entirely through voice commands. By processing natural speech, analyzing the underlying structure of web pages, and mapping spoken instructions to precise actions, the system transforms traditional navigation into a more intuitive and conversational experience. A critical feature is its ability to resolve ambiguity. When faced with multiple similar elements, such as two buttons labeled “Upload,” the system asks the user to clarify their intent, mirroring real-life communication and ensuring that interactions remain accurate and dependable. In its current prototype, the system focuses primarily on voice interaction and real-time understanding of the page. Spoken requests are converted into structured commands by a parser, while the page is broken down through DOM analysis to identify the most relevant interactive elements. This enables support for common actions, such as opening websites, clicking buttons and links, scrolling, navigating menus, and typing into fields, without requiring any site-specific setup. Initial testing shows strong performance on navigation-centric browsing, with an overall success rate of around 80%, driven primarily by robust execution of general navigation and clicking actions (e.g., opening links, selecting buttons, and scrolling). The main weaknesses appear in form-heavy scenarios, especially when entering emails or other structured inputs, where transcription and formatting errors can reduce precision. These results highlight both the practicality of voice-first browsing today and the areas where more intelligent input handling and stricter validation would further enhance reliability. By demonstrating the capabilities of voice-based web navigation, this project showcases how accessibility and modern technology can collaborate to enhance everyday life. It offers not just a practical tool, but also a glimpse of a future where interacting with the web feels more natural, inclusive, and human.

Pages: 89 to 94

Copyright: Copyright (c) IARIA, 2026

Publication date: May 24, 2026

Published in: conference

ISSN: 2308-4138

ISBN: 978-1-68558-383-5

Location: Venice, Italy

Dates: from May 24, 2026 to May 28, 2026