eTALK embodied Thought for Abstract Language Knowledge

语言是人与人之间最自然的交流方式，也是人与机器人互动的最自然方式。由于人们使用的词绝大多数是抽象的词，为了实现与机器人的自然语言交互，机器既能处理具体的概念，也能处理抽象的概念，这是至关重要的。这个时间项目战略性地建立在发展心理学和体现认知理论的抽象概念、发展机器人学和认知建模的人工智能方法以及人与机器人交互和语言使用的最新实质性进展的基础上，以引导我们对机器人代理中抽象概念和单词的基础的科学技术理解。该项目将开发和测试一种新颖的认知发展架构，利用最新的生成性深度学习模型和开发的机器人学方法进行人-机器人交互实验，通过迁移学习具体和抽象的单词。三个案例研究已经被战略性地规划，以展示所提出的方法的广度和稳健性，并证明计算方法在处理各种类型的抽象概念时的灵活性。案例研究的重点是：(I)知觉、运动和情感概念的抽象词基础；(Ii)数字词的学习；(Iii)指示介词、时间词和模糊量词的虚词。这个项目将导致我们在深入理解语言和认知的基础和体现基础的计算方法方面取得突破，并在类人机器人的设计方面取得突破。人形机器人可以学习理解各种抽象单词的对话，并对其采取行动，因为这些对话本质上是基于它们自己的感觉运动经验。该项目将受益于与应用领域的用户和利益相关者的直接接触，以及来自学术界和工业界的国际顾问。

Language is the most natural means of communication among people, as well as for interaction betweenpeople and robots. As the great majority of words people use are abstract words, to achieve natural languageinteraction with robots it is crucial that machines can handle both concrete and abstract concepts. This timelyproject strategically builds on the recent, substantial advances in developmental psychology and embodiedcognition theories on abstract concepts, on developmental robotics and AI methods for cognitive modelling,and in human-robot interaction and language use, to bootstrap our scientific and technological understandingof the grounding of abstract concepts and words in robotic agents. The project will develop and test a novelcognitive developmental architecture utilising the latest generative deep learning models and developmentalrobotics methods for human-robot interaction experiments on the learning of concrete and abstract words viagrounding transfer. Three case studies have been strategically planned to show the breadth and robustness ofthe proposed approach, and to demonstrate the flexibility of the computational approach in dealing withvarious types of abstract concepts. The case studies focus on: (i) abstract word grounding for perceptual,motor and affective concepts, (ii) number word learning, and (iii) function words for demonstrative (deictic)prepositions, words about time and fuzzy quantifiers. This project will lead to breakthroughs in our deeperunderstanding of the computational approaches to the grounding and embodiment bases of language andcognition and in the design of humanoid robots which can learn to understand, and act on, conversationsabout a variety of abstract words, as these are intrinsically grounded in their own sensorimotor experience.The project will benefit from direct engagement with users and stakeholders of the application domains, aswell as from the international advisors from both academia and industry.