EAGER: Persuasive Sensing Networks: A New Frontier to Changing Human Behavior

这个项目探索了一个名为说服感知的新前沿，它将传感器网络和说服技术这两个领域的进展结合在一起。今天，传感器网络的进步使捕获、检测和分析数据成为可能。然而，缺少的是向受试者呈现从传感器数据中挖掘的关于他们的日常生活和活动节奏的相关信息，并使用反馈机制来改变人类的行为。现在证明，人类通常遵循一种大约24小时波动的节律，称为昼夜活动节律。这是一个探索性的研究项目，旨在设计和设计这样一个原型系统。从环境传感器和可穿戴传感器获得的数据被融合在一起，产生有意义的反馈，以说服最终用户。这项研究项目在几个方面都是新颖的。首先，将探测空间(或位置)活动的环境传感器数据与收集生理健康数据并利用两者来检测昼夜活动节奏的可穿戴式传感器融合在一起的想法以前还没有做过。其次，尽管传感器网络以前曾被用于医疗保健应用，但我们基于挖掘数据并将其与正常活动进行基准比较以及检测识别疾病或病理发作的模式的能力的说服性反馈是新颖的。第三，人工神经网络数据挖掘算法可以从昼夜活动节律中识别模式，然后将这些模式与已知节律数据库进行匹配，这将是一项重大的创新贡献。最后，不同的人在接受建议时有很大的不同。为了实现有效性，该系统需要监测和学习人类对先前建议的反应并进行调整。更广泛的影响包括使用无线传感器网络以及有说服力的技术设计，这将为预防和帮助解决肥胖和糖尿病等慢性疾病开辟新的可能性。该项目雇用博士后和研究生，将培训和教育下一代医疗保健IT劳动力，并将研究成果整合到课堂上。这项研究的发现将被纳入克莱蒙特研究生院的研究生课程。

This project explores a new frontier called Persuasive Sensing that brings together advances happening in two fields, namely sensor networks and persuasive technology. Today, advances in sensor networks are making it possible to capture, detect, and analyze data. However what is missing is to present relevant information mined from sensor data to subjects about their daily life and activity rhythms and using feedback mechanisms to alter human behavior. It is now demonstrated that human beings normally follow an approximately 24-hour fluctuating rhythm known as circadian activity rhythm. This is an exploratory research project to design and engineer such a prototype system. The data obtained from environmental sensors as well as body-wearable sensors are fused together to generate meaningful feedback to persuade end-users.This research project is novel in several ways. First, the idea to fuse environmental sensor data that detects activity in space (or location) along with body-wearable sensors that collects physiological health data and utilize both to detect circadian activity rhythms has not be done before. Second, while sensor networks have been used in healthcare applications before, our persuasive feedback based on mining the data and benchmarking it against normal activity and the ability to detect patterns that identify onset of diseases or pathologies is novel. Third, the artificial neural network data mining algorithms that can identify patterns from circadian activity rhythms and then match those against a database of known rhythms will be a significant novel contribution. Finally, people are very different when taking suggestions. In order to achieve effectiveness, the system needs to monitor and learn from human reactions from previous suggestions and adapt. Broader impact includes the use of wireless sensor networks along with persuasive technology design that will open up new possibilities for prevention and help address chronic diseases such as obesity and diabetes. Employing post docs and graduate students, this project will train and educate the next generation of workforce in Healthcare IT and integrate research findings into classroom. Findings from the research will be incorporated into a graduate level course at Claremont Graduate University.