SBIR Phase I: Create Effective COVID-19 Chatbots through High Patient Engagement Phase I

小型企业创新研究(SBIR)第一阶段项目的更广泛影响/商业潜力是提供按需定制的健康信息。聊天机器人是一种自动化沟通代理，在其他领域取得了成功，但在新冠肺炎疫情期间，它们在患者参与方面取得的成功有限。患者参与度的提高会带来更好的健康结果和成本节约。该项目将开发一个新冠肺炎聊天机器人来解决常见的问题和交易，让医护人员能够处理更复杂的病例，并为寻求在社交距离期间尽量减少办公室访问的患者提供信息。这个小型企业创新研究(SBIR)第一阶段项目将通过聊天机器人促进患者参与的发展。这个项目将聊天机器人表示为状态图，其中转换是基于用户消息的。所提出的聊天机器人优化器通过应用强化学习来动态计算给定用户简档和上下文的最佳响应，从而扩展了该模型。优化器的目标是达到成功状态(例如，与咨询师预约)，考虑到用户(患者)可能随时退出的事实；以前针对目标导向的聊天机器人的工作通常不包括这种可能性。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact /commercial potential of this Small Business Innovation Research (SBIR) Phase I project is to provide on-demand customized health information. Chatbots are automated communication agents that have been successful in other sectors, but to date they have enjoyed limited success in patient engagement during the COVID-19 pandemic. Increased patient engagement leads to better health outcomes and cost savings. This project will develop a COVID-19 chatbot to address common questions and transactions, freeing healthcare staff for more complex cases and providing information to patients seeking to minimize office visits during periods of social distancing. This Small Business Innovation Research (SBIR) Phase I project will advance development of patient engagement via chatbots. This project represents a chatbot as a graph of states, wherein a transition is based on the user message. The proposed chatbot optimizer extends this model by applying reinforcement learning to dynamically compute the best response given the user profile and context. The objective of the optimizer is to reach a success state (for example, make an appointment with a counselor), accounting for the fact that a user (patient) may drop out at any time; previous work on goal-oriented chatbots has generally not included this possibility.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.