Programme
Monday 4th (half day)
14:30 - 15:00 Welcome and Introductory remarks
15:00 - 16:00 Invited talk: Verena Rieser "Responsible Conversational AI: Trusted, Safe and Bias-free"
16:00 - 16:15 Short break
16:15-16:45 Angus Addlesee and Arash Eshghi "Incremental Graph-Based Semantics and Reasoning for Conversational AI" (short paper)
16:45 - 17:15 Mauricio Mazuecos, Patrick Blackburn and Luciana Benotti "The impact of answers in referential visual dialog" (short paper)
Tuesday 5th
10:30 - 11:15 Staffan Larsson and Jean-Philippe Bernardy "Semantic classification and learning using a linear tranformation model in a Probabilistic Type Theory with Records" (long paper)
11:15 - 12:00 Jorge Del-Bosque-Trevino, Julian Hough and Matthew Purver - Communicative Grounding of Analogical Explanations in Dialogue: A Corpus Study of Conversational Management Acts and Statistical Sequence Models for Tutoring through Analogy (student paper)
12:00 - 13:15 Lunch break
13:15 - 14:00 Katie Baker, Amit Parekh, Adrien Fabre, Angus Addlesee, Ruben Kruiper and Oliver Lemon "The Spoon is in the Sink: Assisting Visually Impaired People in the Kitchen" (student paper)
14:00 - 14:45 Ellen Breitholtz and Robin Cooper "Dogwhistles as Inferences in Interaction" (long paper)
14:45 - 15:15 Simeon Schüz and Sina Zarrieß "Decoupling Pragmatics: Discriminative Decoding for Referring Expression Generation" (short paper)
15:15 - 15:30 Short break
15:30 - 16:30 Invited talk: Invited Speaker - Joyce Chai "Language, Action, and Physical Causality"
Wednesday 6th
10:00 - 11:00 Invited talk: Riccardo Fusaroli "Interactive linguistic alignment in perspective: how do we re-use each other’s language to understand each other, solve problem together and learn language"
11:00 - 11:15 Short break
11:15 - 11:45 Georgiy Platonov, Benjamin Kane and Lenhart Schubert "Generating Justifications in a Spatial Question-Answering Dialogue System for a Blocks World" (short paper)
11:45 - 12:30 Vladislav Maraev, Ellen Breitholtz, Christine Howes and Jean-Philippe Bernardy "Why should I turn left? Towards active explainability for spoken dialogue systems" (student paper)
12:30 - 13:30 Lunch break
13:30 - 15:15 NLI in Dialogue: presentation of the dataset and preliminary results
Invited Speakers
Verena Rieser, Heriot-Watt University, Edinburgh
Responsible Conversational AI: Trusted, Safe and Bias-free
With recent progress in deep learning, there has been an increased interest in learning dialogue systems from data, also known as “Conversational AI”. In this talk, I will focus on the task of response generation, for which I will highlight lessons learnt and ongoing challenges, such as reducing `hallucinations’ for task-based systems, safety critical issues for open-domain chatbots, and the often overlooked problem of `good’ persona design. I will argue that we will need to solve these challenges to create trusted, safe and bias-free systems for end-user applications.
Joyce Chai, University of Michigan
Language, Action, and Physical Causality
Language communication plays an important role in human learning and skill acquisition. With the emergence of a new generation of cognitive robots, it becomes increasingly important to empower these embodied agents with an ability to learn directly from human partners through language. One of the key challenges is to connect language with actions in the physical world, whether it is to perceive actions or to execute actions. While we humans develop an understanding of physical actions and their cause and effect to the physical world at a very young age, such understanding of physical causality is a major bottleneck in machine intelligence. In this talk, I will give a brief introduction to our recent work on grounding language to actions in the context of communicative task learning. I will highlight the important role of physical causality in this process and discuss key research challenges and opportunities.
Riccardo Fusaroli, Aarhus University
Interactive linguistic alignment in perspective: how do we re-use each other’s language to understand each other, solve problem together and learn language
In this talk I will present an overview on interactive linguistic alignment, the tendency to re-use one’s interlocutor’s linguistic forms. From an early enthusiastic foray (Fusaroli et al 2012, Coming to terms), I’ll move onto the challenges of building standardized tools to analyze alignment (Duran et al 2019, ALIGN) and the overview that is emerging when applying these tools to a variety of domains: diverse conversational tasks, diverse languages, diverse language acquisition stages and clinical populations.
Accepted papers
Student papers
The Spoon is in the Sink: Assisting Visually Impaired People in the Kitchen
Katie Baker, Amit Parekh, Adrien Fabre, Angus Addlesee, Ruben Kruiper and Oliver Lemon
Communicative Grounding of Analogical Explanations in Dialogue: A Corpus Study of Conversational Management Acts and Statistical Sequence Models for Tutoring through Analogy
Jorge Del-Bosque-Trevino, Julian Hough and Matthew Purver
Why should I turn left? Towards active explainability for spoken dialogue systems.
Vladislav Maraev, Ellen Breitholtz, Christine Howes and Jean-Philippe Bernardy
Short papers
Incremental Graph-Based Semantics and Reasoning for Conversational AI
Angus Addlesee and Arash Eshghi
The impact of answers in referential visual dialog
Mauricio Mazuecos, Patrick Blackburn and Luciana Benotti
Generating Justifications in a Spatial Question-Answering Dialogue System for a Blocks World
Georgiy Platonov, Benjamin Kane and Lenhart Schubert
Decoupling Pragmatics: Discriminative Decoding for Referring Expression Generation
Simeon Schüz and Sina Zarrieß
Long papers
Dogwhistles as Inferences in Interaction
Ellen Breitholz and Robin Cooper
Semantic classification and learning using a linear tranformation model in a Probabilistic Type Theory with Records
Staffan Larsson and Jean-Philippe Bernardy