PhD in Computing Science
Table of Contents
Tue Apr 16, 2019, from 10ā11am MST, in the Computing Science Centre (room 3-33) at the University of Alberta, I presented a public presentation of my thesis work. You can watch the recorded live stream here (slides to follow along with):
Following the talk, I was sequestered to a private examination room. In the room, the thesis committee (including supervisor Dr. Patrick M. Pilarski) asked questions, made comments, then decided that my thesis, the oral presentation, and the examination were worthy of granting me a doctorate of philosophy (Ph.D.) in computing science.
My responsibility was to demonstrate to the satisfaction of the examining committee that I possess: 1) adequate knowledge of the thesis and an ability to defend it, and, more generally, 2) the ability to pursue and complete original research at an advanced level.
Humour-in-the-loop: Improvised Theatre with Interactive Machine Learning Systems #
Improvisation is a form of live theatre where artists perform “real-time, dynamic problem solving” to collaboratively generate interesting narratives (Magerko et al., 2009).
The main contribution of my thesis is the development of artificial improvisation: improvised theatre performed by humans alongside intelligent machines.
In this talk, I start by presenting the background underlying the art of improvisation and the scientific fields of interactive machine learning and dialogue generation.
Then, I present Pyggy, the first experiment on live stage human-machine improvisation and A.L.Ex., the Artificial Language Experiment (from HumanMachine) which addresses several critical technical improvements over Pyggy. Improbotics is then presented which details audience evaluation of Turing test-inspired live improvised performance using A.L.Ex. (Mathewson and Mirowski, 2018).
Two novel contributions to machine-assisted narrative generation are then presented and discussed. The first of these contributions, Shaping the Narrative Arc, is a model incorporating an underlying narrative arc to improve response generation (Mathewson et al., 2019). The second contribution, dAIrector, synthesizes a plot graph with contextual information to generate contextual plot points and serve as the director (Eger and Mathewson, 2018).
The talk concludes by discussing public reflections on live artificial improvisation performances from around the world and interesting future directions to explore – including automated presentation generation (Winters and Mathewson, 2019) and improved automated dialogue evaluation metrics (Dziri et al., 2018, 2019).
My work presents fundamental advances in human-machine interaction through the lens of improvised theatre, the ideal evaluation testbed for collaboration between humans and intelligent machines.
This work would not be possible without a huge amount of support from collaborators around the world.