Friday, 3 December, 2021
QUT Gardens Point Campus & Online
The Responsible Data Science Symposium aims to bring together:
- Academic experts to deliver presentations on key "responsible data science" topics from across disciplines (including statistics, computer science, ethics, law, media, and business)
- Practitioners from the ABS, Defence, and other Australian Public Service organisations, to present on issues, challenges, opportunities, and proposed projects
- Invited guests, including students, researchers, and practitioners, from across sectors (academia, research, industry, and government) interested in engaging in discussions and workshops, to share and advance knowledge, develop relationships, and identify and progress opportunities
Some attendees may also want to be involved in
- ABS research projects (outlined below), or others, explored in the workshops, and/or
- collaborating to create academic/practitioner outputs from the workshops;
and the Symposium aims to support the advancement of such opportunities.
QUT Hosted Hybrid Event
The event will be hosted at QUT (for in-person and virtual attendance), with full catering provided for attendees.
Preliminary Symposium Program
The Symposium will run from 9 am-5 pm - with catered breaks (for morning tea, lunch, and afternoon tea) plus the option to enjoy a celebratory drink and networking on conclusion. The Symposium will open for registrations from 8:15am.
Overview of Day: (times are AEST - Brisbane)
8:15 am - 9:00 am: Pre-Symposium Registration opens for Attendees
9:00 am - 5:00 pm: Symposium (featuring keynote and other presentations, break-out discussion groups, workshops, opportunity identification, and closing remarks)
5:00 pm - 6:00pm: Celebratory drink and Networking
On the eve of the Symposium, a Pre-Symposium Dinner will be hosted for guest speakers, partners and invited guest to attend.
Opportunities including Projects and Publications
The "responsible data science" theme and sub-topics including ethical AI, explainable AI, and uncertainty quantification, are opportunity rich from a knowledge advancement and collaboration perspective.
The symposium will enable some exploration of these topics, opportunities, and relationship-building for future collaborations.
ABS Research Projects:
The QUT Centre for Data Science and ABS have committed to the following projects, to be explored at the Symposium, and aim to form project teams (with opportunities for other experts, students, RAs, and supervisors to join) for project commencement in 2021/2022:
- Investigate the use of attention mechanisms for explainability;
- Estimating uncertainty for Deep Neural Networks (DNNs) (building on earlier work);
- Investigate applications of contrastive explanations in the ABS;
- Efficient Sampling Techniques for Labelling Training Data for Model Building; and
- Representivity of Training and Test Data;
and others, including new opportunities which may be identified from the workshop.