Since the inception of human research studies, researchers must often interact with participants on a set schedule to collect data. Researchers manually perform many interactions, leading to considerable time and financial expenses. Usually, user-provided data collection consists of surveys administered via telephone or email. These methods are tedious for the survey administrators, which could cause fatigue and potentially lead to collection mistakes. This project leverages recent advancements in automatic speech recognition, speech-to-text, natural language understanding (NLU), and finite-state machines to automate research protocols. This generalized application is fully customizable and irrespective of any research study. New research protocols can be quickly created based on these parameters once envisioned. Thus, we present SmartState, a fully-customizable, state-driven protocol manager combined with supporting AI components to autonomously manage user data and intelligently determine users' intentions through chat and end-device interactions.

翻译：自人类研究启动以来，研究人员通常需按预定时间表与参与者进行交互以收集数据。这类交互多依赖人工操作，导致大量时间与财务成本。传统用户数据收集常采用电话或电子邮件发放问卷的方式，调查管理人员执行此类操作时易感疲劳，甚至可能引发数据采集错误。本项目利用自动语音识别、语音转文本、自然语言理解及有限状态机等前沿技术，实现研究协议的自动化。该通用化应用完全可定制，不受具体研究领域限制。一旦构想出新的研究方案，即可基于既定参数快速创建对应研究协议。由此，我们提出SmartState——一种结合人工智能组件的、完全可定制的状态驱动协议管理器，能够通过聊天和终端设备交互自主管理用户数据并智能识别人机意图。