Dr John Taylor

Overview

Recommender systems are everywhere - you can find them on websites suggesting you buy certain socks; they are used to help filter and present relevant content in large libraries and databases; you can even find them in advanced computer and robotics systems in fields such as health and surgical medicine, defence, and aerospace.

But not all recommendations are useful to the operator. It is not always clear why a human should trust a recommendation, or why they should accept some recommendations over others. As AI and machines play an increasingly important role in our lives, how can we make their recommendations appear logical and serendipitous to the end-user?

Dr Taylor is a behavioural research scientist in human machine interaction and human autonomy teaming. He researches how to tap into the psychology of the end-user to predict what they need so a machine can present good recommendations that the human will accept and trust.

By incorporating human factors such as psychology, perception and cognitive load, it is possible to improve the technology by looking at how the machine provides information or recommendations to the user.

Dr Taylor’s approach combines his knowledge of recommender systems with empirical psychology research that investigates the human factors that influence human machine interaction.

While most recommender systems use existing data to recommend new items or actions to users, Dr Taylor’s research is different because he incorporates into his recommender systems perception and how humans interact with machines.

The ultimate aim of this work aims to enhance trust and performance so that the human and the machine become a more effective team, for any given task, irrespective of its duration.

Impact

Dr Taylor has developed numerous recommender systems for use in interactive media, robotics and industry.

He has developed recommender systems in partnership with academia and the creative arts industry (Australian Music Centre) to promote Australian music; cognitive-based human machine interfaces for the Australian Government Defence Cooperative Research Centre For Trusted Autonomous Systems (TAS-CRC); and interactive and intelligent creative systems that have been exhibited at both national and international arts festivals.

His research has also contributed systems that can be remotely controlled and monitored, to replace operators that previously worked in dangerous environments.

Why it matters

Dr Taylor’s research is significant to all industries where there is human-machine interaction. It has applications, for example, for user interface design and visual perception; auditory perception and operator warnings; cognitive load and whether a machine can adapt to a user that is struggling with a task; hardware design; training in autonomous systems; information processing, analysis, and presentation.

An interesting aspect of his research is human/machine interaction that lasts for several hours. His research has shown how the situation and task evolve over longer periods of time, and has relevance to industries such as advanced manufacturing, or long-haul flights, where an operator may monitor a system for extended periods and yet be expected to make optimal decisions in critical environments.

Ultimately, he hopes to develop recommender systems that can help save lives by parsing a lot of information more quickly than a human and help in human decision-making.

I imagine a future where robots, AI, and autonomous systems are used for tasks that would be otherwise impossible for a human, but one in which the human has ultimate control.

Career

Dr Taylor was an industry professional and also an expert in auditory perception. A certified UX tester and designer with the British Computer Society, his passion for computers quickly became a passion for understanding the neural mechanisms of how humans and machines interact.

He takes an interdisciplinary and collaborative approach to developing trusted recommender systems, but the most important aspect of his approach is in learning the requirements and domain knowledge of the end-user.

Collaboration

Dr Taylor has collaborated widely throughout his career. For example, he contributed to the development of autonomous systems for mine counter measures, which involved multiple universities and industry partners.

He welcomes collaborations with the aerospace, defence, music and art, transportation, disaster relief, health and manufacturing industries.

Connect