Recently, there has been an increase in industrial and academic research on data-driven analytics with smartphones based on the collection of app usage patterns and surrounding context data. The Android mobile operating system utilizes Usage Statistics API (US API) and Accessibility Service API (AS API) as representative APIs to passively collect app usage data. These APIs are used for various research purposes as they can collect app usage patterns (e.g., app status, usage time, app name, user interaction state, and smartphone use state) and fine-grained data (e.g., user interface elements \& hierarchy and user interaction type \& target \& time) of each application. In addition, other sensing APIs help to collect the user's surroundings context (location, network, ambient environment) and device state data, along with AS/US API. In this review, we provide insights on the types of mobile usage and sensor data that can be collected for each research purpose by considering Android built-in APIs and sensors (AS/US API, and other sensing APIs). Moreover, we classify the research purposes of the surveyed papers into four categories and 17 sub-categories, and create a hierarchical structure for data classification, comprising three layers. We present the important trends in the usage of Android's built-in APIs and sensors, including AS/US API, the types of data collected using the presented APIs, and discuss the utilization of mobile usage and sensor data in future research.
翻译:最近,在收集应用程序使用模式和周围环境数据的基础上,用智能手机进行数据驱动分析的工业和学术研究有所增加;Android移动操作系统使用使用用户统计数据API(US API)和无障碍服务API(AS API)作为具有代表性的API来被动收集应用程序使用数据;这些API用于各种研究目的,因为它们能够收集应用程序使用模式(例如应用程序状况、使用时间、应用程序名称、用户互动状态和智能手机使用状态)和每种应用的精细数据(例如用户界面元素 等级和用户互动使用类型 目标 + 时间);此外,其他感测API帮助收集用户周围的环境(地点、网络、环境环境)和装置状态数据,同时收集AS/US API。 在本次审查中,我们通过考虑API和传感器的固定用途和传感器(AS/US)中的用户界面和用户互动使用类型 目标 + 目标 时间 目标 时间 ;此外,其他感测API 帮助收集用户周围的环境环境(地点、网络、环境环境)和设备状态数据数据。