Automated Detection of TalkBack Interactive Accessibility Failures in Android Applications
Automated Detection of TalkBack Interactive Accessibility Failures in Android Applications
复制标题
自动检测 Android 应用程序中的 TalkBack 交互式辅助功能故障
DOI：
10.1109/icst53961.2022.00033
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
William G. J. Halfond
中科院分区：
文献类型：
--
作者：
Ali S. Alotaibi;Paul T. Chiou;William G. J. Halfond
TalkBack是全球3.04亿视障用户使用安卓设备时最受欢迎的屏幕阅读器之一。然而，许多流行的安卓应用缺乏适当的TalkBack无障碍访问机制，这可能导致应用无法使用的故障。现有的与无障碍访问相关的技术在检测这些问题方面存在局限性。在本文中，我们提出了一种新的方法，用于自动检测TalkBack交互无障碍访问故障，这些故障会阻止TalkBack用户与安卓应用的核心功能进行交互。我们在实际安卓应用上对我们的技术进行评估，结果表明它能够高精度且高召回率地检测这些TalkBack无障碍访问故障。
TalkBack is one of the most popular screen readers for the 304 million visually impaired users worldwide to access Android devices. Yet, many popular Android apps lack a proper TalkBack accessible mechanism, which could cause failures that make apps unusable. Existing accessibility related techniques are limited in terms of detecting these issues. In this paper, we present a novel approach for automatically detecting TalkBack interactive accessibility failures that prevent TalkBack users from interacting with core functionalities in Android apps. Our evaluation of our technique on real-world Android apps showed that it is able to detect these TalkBack accessibility failures with high precision and recall.