Student Participant Support for Conversational Intelligence Summer School 2019

从商业个人助理和客户服务聊天机器人到智能家教和语音控制界面，对话系统和对话代理目前在各个行业都有很高的需求，并且在我们的日常生活中无处不在。虽然人工智能的最新进展使这些实际应用变得触手可及，但在这些系统能够支持人类级别的对话之前，还需要解决一些严峻的挑战。对话智能暑期学校（CISS）旨在培养对这一具有挑战性和令人兴奋的领域感兴趣的年轻研究人员，并为他们提供开发此类应用的动手技能。这个为期一周的强化课程为参与者提供了建立对核心会话AI概念和技术的坚实理解的机会，目标是培养一批有才华的实践者，以解决该领域未解决的问题。本提案要求资助学生参加将于2019年6月24日至29日在马萨诸塞大学洛厄尔分校举行的第二届国际会话智能暑期学校（CISS 2019）。参与者将接受有关开发聊天会话代理的当代深度学习技术的实践培训，提供(a)了解会话代理人工智能的当前研究趋势，他们将能够在自己的研究中使用；(b)行业需求的实用技能，他们将能够将其运用到劳动力中。该计划包括讲座指导、编程教程、顶级研究人员的邀请讲座和监督项目工作。参与的学生将有机会与其他国内和国际机构的同龄人、导师和资深组织者互动。在客座讲座之后的几个小时里，学生们可以与受邀进行客座讲座的知名研究人员进行密切的互动。该项目汇集了来自不同国家的学生和研究人员，使参与者接触到更广泛的教育和研究视角，并建立有助于促进未来国际合作的联系。该项目提供的机会鼓励学生参与者继续努力发展技术专长，并进一步提高他们对人工智能研究前沿职业的兴趣。项目代码、教材和邀请的讲座将尽可能公开，以增加该项目在直接参与者受众之外的影响。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

From commercial personal assistants and customer service chatbots to intelligent tutors and voice control interfaces, dialogue systems and conversational agents are presently in high demand in a variety of industries and are becoming ubiquitous in our everyday lives. While recent advances in artificial intelligence have placed some of these practical applications within reach, some tough challenges need be addressed before such systems are able to support human-level conversation. The Conversational Intelligence Summer School (CISS) aims to train young researchers interested in this challenging and exciting field in state-of-the-art techniques and provide them with hands-on skills to develop such applications. This one-week intensive program offers participants the opportunity to build a solid understanding of core conversational AI concepts and techniques, with the goal to train a core of talented practitioners to take on unsolved problems in this domain. This proposal requests funding to subsidize student participation in the 2nd International Summer School for Conversational Intelligence (CISS 2019), to be held at the University of Massachusetts Lowell from June 24 to 29, 2019. Participants will receive hands-on training in contemporary deep learning techniques for developing chit-chat conversational agents, providing (a) exposure to current research trends in AI for conversational agents, which they will be able to use in their own research and (b) practical skills in demand in the industry which they will be able to carry on to the workforce. The program includes a combination of lecture instruction, programming tutorials, invited lectures by top researchers, and supervised project work. Participating students will benefit from the opportunity for interaction with their peers at other domestic and international institutions, as well as instructors and senior organizers. Networking hours following guest lectures allow the students to have close interactions with prominent researchers invited to give guest lectures. The program brings together students and researchers from different countries, exposing participants to a wider range of educational and research perspectives and creating ties that can help to foster future international collaborations. The opportunities provided by this program encourage student participants to continue in their efforts to develop technical expertise and further their interest in careers at the forefront of AI research. Project code, teaching materials, and invited lectures will be made accessible publicly whenever possible in order to increase the impact of this program outside of the audience of immediate participants.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.