The paper deals with the design of an agent which modifies and enhances the various alert systems in the smartphones. The actions of the agent includes sorting the notifications abiding to human thinking, helping the user to have a safe conversation, assisting in tracking back the reach-ability status of the caller when needed, conveying the user about the notifications in times of situations like drained battery and smartly alerting the user in situations like sleeping. The agent uses the information gathered from a survey, to modify the existing methods of alerts and produce alerts which abide by the human cognitive responses.
CovidAlert -- A Wristwatch-based System to Alert Users from Face Touching
Northern Illinois University(北伊利诺伊大学)
全球2019年有2019万人感染,450万人在持续的新冠肺炎大流行中丧生。在疫苗广泛使用之前,预防措施和安全措施,如戴口罩,身体距离,避免面部接触是遏制病毒传播的主要手段。摸脸是一种强迫性的人类行为,如果没有持续的有意识的努力,就无法阻止,即使这样,它也是不可避免的。为了解决这个问题,我们设计了一个基于智能手表的解决方案,CovidAlert,它利用随机森林算法训练来自智能手表的加速度计和陀螺仪数据,以检测手到脸的过渡,并向用户发送快速触觉警报。CovidALert具有高能效,因为它采用STA/LTA算法作为看门人,在用户不活动时减少手表上随机森林模型的使用。我们的系统的总体准确率为88.4%,假阴性和假阳性率较低。我们还通过在商业Fossil Gen 5智能手表上实施该系统来证明其可行性。
Actionable Cybersecurity Notifications for Smart Homes: A User Study on the Role of Length and Complexity
Leipzig University(莱比锡大学);Center for Scalable Data Analytics and Artificial Intelligence (ScaDS.AI) Dresden/Leipzig(德累斯顿/莱比锡可扩展数据分析与人工智能中心);University of Kassel(卡塞尔大学)