The following is a preliminary programme. Paper slots are subject to change depending on author availability.
All talks will take place in the Paccar Theatre in Science Gallery Dublin
* Denotes the listed presenters.
Day 1 - Thursday 22nd August
08:15 - 09:00 Registration
---------------------------------------------------------------------------------------------------------------------
09:00 - 09:10 Welcome to CUI 2019
Benjamin Cowan & Leigh Clark (CUI 2019 General Chairs)
---------------------------------------------------------------------------------------------------------------------
09:10 - 10:15 Keynote 1: Cosmin Munteanu (University of Toronto Mississauga)
My dearest Alexa, would you be so kind as to let me know what the bloody weather may be today?
Oscar Wilde once said: “Conversation about the weather is the last refuge of the unimaginative.” This suggests that, for more than four decades, the design and development of interfaces that allow humans to interact more naturally with machines had a significant creativity deficit. While in the early days of dialogue systems and interactive voice responsive applications this could have been largely attributed to engineering limitations, our current expectations and needs challenge our views of what (speech-based) conversational interfaces can and should do for us – beside informing us about the weather. In this talk, I argue that, while currently available interfaces do indeed demonstrate technological progress, inertia in both design and engineering efforts has lead to interfaces that have yet to reach the holy grail: seamless interactions with new media, information, and devices, through natural modalities that transcend the confines of conventional interfaces. The consequences of this are particularly evident within contexts where natural, conversational interfaces may help marginalized users that stand to benefit the most from such technologies, yet fail to sustainably do so. Grounded in more than two decades of research in this space, I deconstruct in this talk some of the claims and myths about spoken conversational interfaces that hold us back from designing truly useful applications. I then discuss how we may approach the design of such interfaces in ways that are not only more imaginative, but that will also lead to the emergence of interactions that are more meaningful and more inclusive to a broader range of users and contexts.
10:15 - 11:10 Coffee Break
---------------------------------------------------------------------------------------------------------------------
11:10 - 12:00 Paper Session 1: Chatbots
Chair: Joel Fischer
11:10 - 11:30 Typefaces and the Perception of Humanness in Natural Language Chatbots
Heloisa Candello*, Claudio Pinhanez, Flavio Figueiredo
11:30 - 11:40 Chatbots as Unwitting Actors
Allison Perrone, Justin Edwards*
11:40 - 12:00 Can Direct Address Affect User Engagement with Chatbots Embodied in Physical Spaces?
Heloisa Candello*, Claudio Pinhanez, Mauro Pichiliani, Marisa Vasconcelos, Haylla Conde
---------------------------------------------------------------------------------------------------------------------
12:00 - 13:00 Lunch
---------------------------------------------------------------------------------------------------------------------
13:00 - 14:30 Paper Session 2: IPA Use
Chair: Mary Ellen Foster
13:00 - 13:20 Patient and Consumer Safety Risks When Using Conversational Assistants for Medical Information: An Observational Study of Siri, Alexa, and Google Assistant
Timothy Bickmore*, Ha Trinh, Stefan Olafsson, Teresa O'Leary, Reza Asadi,
Nathaniel Rickles, Ricardo Cruz
13:20 - 13:40 Voice Interfaces in Everyday Life
Martin Porcheron*, Joel Fischer, Stuart Reeves, Sarah Sharples
13:40 - 14:00 Multitasking with Alexa: How Using Intelligent Personal Assistants Impacts Language- based Primary Task Performance
Justin Edwards*, He Liu Liu, Tianyu Zhou, Sandy Gould, Leigh Clark, Phillip Doyle,
Benjamin R. Cowan
14:00 - 14:10 Finding Contextual Meaning of the Wake Word
Hyunhoon Jung*, Hyeji Kim
14:10 - 14:20 Inquisitive Mind: A Conversational News Companion
Mateusz Dubiel*, Alessandra Cervone, Giuseppe Riccardi
14:20 - 14:30 Voice Assistants and Older People: Some Open Issues
Sergio Sayago*, Barbara Barbosa, Benjamin R. Cowan
---------------------------------------------------------------------------------------------------------------------
14:30 - 15:00 Coffee Break
---------------------------------------------------------------------------------------------------------------------
15:00 - 16:10 Paper Session 3: Dialogue & Conversational Principles
Chair: Susan E. Brennan
15:00 - 15:20 Explorations in multiparty casual social talk and its relevance for social human machine dialogue
Emer Gilmartin*, Benjamin R. Cowan, Carl Vogel, Nick Campbell
15:20 - 15:30 I Don’t Know What You’re Talking About, HALexa
Christine Murad*, Cosmin Muntenau
15:30 - 15:40 Conversational design considered harmful?
Stuart Reeves*
15:40 - 15:50 How language works and what machines can do about it
Peter Wallis*, Bruce Edmonds
15:50 - 16:00 Shoehorning in the Name of Science
Jens Edlund*
---------------------------------------------------------------------------------------------------------------------
16:05 - 17:15 Industry Panel
Chair: Cosmin Munteanu
Panelists: Grace Hughes (Fjord @ The Dock), Andrew Ku (Google), Conor Whelan (Voysis)
---------------------------------------------------------------------------------------------------------------------
17:15 - 18:15 Partner Expo & Drinks Reception
---------------------------------------------------------------------------------------------------------------------
20:00 - late Dinner at Drury Buildings
All guests to be seated by 20:00.
Day 2 - Friday 23rd August
09:00 - 10:30 Paper Session 4: System Development & Multimodality
Chair: Sergio Sayago
09:00 - 09:20 Trust in artificial voices: A “congruency effect” of first impressions and behavioural experience
Ilaria Torre*
09:20 - 09:40 Patters of Gaze in Speech Agent Interaction
Razan Jaber*, Donald McMillan, Jordi Solsona Belenguer, Barry Brown
09:40 - 10:00 Crowdsourcing a self-evolving dialog graph
Patrik Jonell*, Per Fallgren, Fethiye Irmak Doğan, José Lopes, Ulme Wennberg,
Gabriel Skantze
10:00 - 10:10 Conversation is Multimodal - Thus Conversational User Interfaces should be as well
Stefan Schaffer*, Norbert Reithinger
10:10 - 10:20 Face-to-Face Conversation: Why Embodiment Matters for Conversational User Interfaces
Mary Ellen Foster*
---------------------------------------------------------------------------------------------------------------------
10:30 - 11:00 Coffee Break
---------------------------------------------------------------------------------------------------------------------
11:00 - 12:10 Paper Session 5: Ethics, Privacy & Trust
Chair: Timothy Bickmore
11:00 - 11:20 “It’s Small Talk, Jim, But Not as We Know It.” Engendering Trust through Human-Agent Conversation in an Autonomous, Self-Driving Car
David R Large, Leigh Clark*, Gary Burnett, Kyle Harrington, Jacob Luton, Peter Thomas, Pete Bennett
11:20 - 11:30 A Need for Trust in Conversational Interface Research
Justin Edwards*, Elaheh Sanoubari
11:30 - 11:40 “I am from all over the world”: Moving towards a healthier voice enabled internet by acknowledging how it is built
Selina Jeanne Sutton*
11:40 - 11:50 From sex and therapy bots to virtual assistants and tutors: How emotional should artificially intelligent agents be?
Stella George*
11:50 - 12:00 Issues Relating to Trust in Care Agents for the Elderly
Brendan Spillane*, Emer Gilmartin, Christian Saam, Vincent Wade
12:00 - 12:10 Who owns your voice? Ethically sourced voices for non-commercial TTS applications
Kristen Scott*, Simone Ashby, David Braude, Matthew Aylett
---------------------------------------------------------------------------------------------------------------------
12:10 - 13:10 Lunch
---------------------------------------------------------------------------------------------------------------------
13:10 - 14:40 Paper Session 6: Voice & Language Design
Chair: Christine Murad
13:10 - 13:30 Siri, Echo and Performance: You have to Suffer Darling
Matthew Aylett*, Benjamin R. Cowan, Leigh Clark
13:30 - 13:50 Progressivity for Voice Interface Design
Joel Fischer*, Stuart Reeves, Martin Porcheron, Rein Sikveland
13:50 - 14:10 What’s in an accent? The impact of accented synthetic speech on lexical choice in human-machine dialogue
Benjamin R. Cowan*, Philip Doyle, Justin Edwards, Diego Garaialde, Ali Hayes-Brady,
Holly P. Branigan, João Cabral, Leigh Clark
14:10 - 14:30 Entertaining and Opinionated but Too Controlling: A Large-Scale User Study of an Open Domain Alexa Prize System
Kevin Bowden*, Jiaqi Wu, Wen Cui, Juraj Juraska, Vrindavan Harrison, Brian Schwarzmann, Nicholas Santer, Stephen Whittaker, Marilyn Walker
14:30 - 14:40 The Right Kind of Unnatural: Designing a Robot Voice
Matthew Aylett*, Selina Jeanne Sutton, Yolanda Vazquez Alvarez
---------------------------------------------------------------------------------------------------------------------
14:40 - 15:10 Coffee Break
---------------------------------------------------------------------------------------------------------------------
15:10 - 16:15 Keynote 2: Susan E. Brennan (The State University of New York at Stony Brook)
Can Machines Learn to Converse? What We Know from >30 Years of "Conversational Interfaces"
Deep learning has revolutionized automated speech recognition. But what about spoken conversation - can machines learn to converse with people? Because models are only as good as the data they're trained on, the answer to that question depends on (at least) two things: what your theory of conversation is, and how well-represented conversational processes are by the artifacts and evidence present in the input data. I will consider conversational processes such as interpersonal coordination, repairs, multimodal cues, the joint construction of meaning, entrainment, and audience design. Just as learning in the domain of image recognition is tainted by unintended biases, learning for spoken dialogue is tainted by a lack of indexical content in the input data. Fortunately, people are adaptable and can cope with many kinds of interactive partners, as long as there is transparency in how the partners present themselves.
---------------------------------------------------------------------------------------------------------------------
16:15 - 16:55 Close & Handover to CUI 2020