Top of the class: UNSW Sydney uses AI to power personalized paths to student success

Applications of AI

UNSW Sydney is one of the world's leading universities, ranked in the top 20 in the QS World University Rankings 2025. This prestigious position is testament to the university's academic excellence and innovative culture.

“There are lots of fantastic innovators at UNSW,” says Simon McIntyre, director of education innovation, “and one of the things we’re working hard to do is bring them together more closely to maximise the impact of their work.”

This culture is driving UNSW to embrace emerging technologies such as AI to improve teaching and learning outcomes – in fact, AI is a key pillar of the UNSW Education Technology Roadmap 2024-2028.

“These technologies will be incredibly important to the future of education because they enable a level of personalization that we've never seen before,” McIntyre says.

Personalize student learning with AI

One way UNSW is using AI is through the Student Learning and Support Data Analytics project, led by the University's Learning Analytics Intelligence team, with significant support and contributions from the IT and UNSW Planning and Performance (UPP) teams. The project uses machine learning to detect early on when students are at risk of academic failure, and connect them with the right support and services when they need it most.

“By quickly identifying those who need help and helping them make the right decisions to seek the help they need, we hope to create a better, more supportive environment for our students,” McIntyre said.

The project employs a modular approach built around the Academic Success Monitor (ASM), a predictive machine learning model trained on historical data from learning and management systems that identifies potential academic risks based on student engagement in the digital learning environment, enabling faculty and students to take proactive steps.

Simon McIntyre, Director of Educational Innovation

The Learning Analytics Intelligence team also worked closely with UNSW’s Student Support Services to develop AI-generated recommendations based on individual students’ circumstances.

“By working with support services we’re able to understand how the language they use, the types of student personas they see, and the syntax they use in their communications escalates at different risk levels,” McIntyre says, “so we’re using this information to create a matrix which we then feed into our AI model.

“We're not replacing support services or working for them – we're just systematising the approach. We make students aware of appropriate support options and give them the freedom to take action independently. We also give support teams 'advance notice' as early as the second week of the term so they can contact students who require more specialised help,” McIntyre says.

ASM for this project leverages a range of Microsoft solutions, including Azure, Azure Machine Learning Studio, Azure OpenAI Service and Power Apps.

Test the benefits

ASM's first small-scale test in 2023 involved 33 faculty members and 25 courses across all faculties based at the University of New South Wales in Sydney. Results were promising, with the model confidently identifying 79% of at-risk students within the first few weeks of the course.

The test was then expanded to an 80-course pilot in early 2024, involving nearly 17,000 students and 83 faculty. ASM identified 284 students at risk of failing and in need of support, and provided faculty with updates and insights into student engagement. Additionally, 75% of faculty said ASM identified potential risks much sooner than before, and 49% of students who received active outreach from the system saw a statistically significant increase in engagement.

Associate Professor Lynne Gribble, from the University of New South Wales' School of Business and Governance, has experienced the benefits of this project first-hand.

“Participating in this project has given me, as an instructor, a quantitative understanding of my students,” she says. “I know that students who leave something to the end are going to disappear from Moodle. [UNSW’s learning management system]or students who are not actively engaged with the course material, will not perform as well as students who are actively engaged.

“can [also] We will personalize ASM messaging to these students and direct them to support services to help them get back on top.”

McIntyre believes this holistic approach is what makes the project unique: “We believe we're one of the first universities to think of this whole thing as a connected ecosystem,” he says.

“We don't put the onus on individual faculty to interpret and act on the data. We make appropriate suggestions based on the data, and we also provide information about student engagement and tailored advice to help them succeed in their own context.”

This project has given our support team greater exposure, reach and insight into a larger group of students than ever before, and we are working to make our services more accessible and targeted to those who could benefit most.”

Enabling responsible use of AI at scale

Undertaking such a comprehensive project was not without challenges. McIntyre says the UNSW team first had to collaborate on developing a more AI-enabled data infrastructure. UNSW is working with Microsoft Industry Solutions Delivery to further explore and prioritize the expansion of AI use cases through its Three Horizon plan, supporting the architectural framework and building the organization's AI aptitude over the long term.

“The data was scattered across the university and not necessarily integrated, so my team worked closely with the UPP and IT teams to build a data lake that allowed us to use AI and ML at scale,” he explains.

McIntyre says collaboration with UNSW Chief Data and Insights Officer (CIDO), Kate Carruthers, and Microsoft partners Accenture and Altis, was key to the project’s success.

“The support from Microsoft and their partner Accenture really helped us get everything started through the joint development of a prototype in their Power Apps Innovation Center program. Then our own CIDO and Altis helped us wire up the custom configuration. [of our Microsoft technology stack] “By working together, we could have done it so quickly on our own,” he says.

Ensuring responsible and ethical use of AI is also a top priority. A steering committee including UNSW students, faculty, education innovation teams and members of the legal department will oversee the project. A privacy impact assessment has also been carried out to ensure compliance with the law, and the University's student privacy agreement has been updated to include the use of AI.

“We have worked extensively with student groups and spoken directly to them about what they are happy with and what they are not happy with. [about the use of AI]and then we collaborate with them to design a solution,” McIntyre says.

Further expansion and consolidation

Looking ahead, UNSW has ambitious plans for its Data Insights Project and related initiatives around student learning and support: ASM will be rolled out to all new students and faculty in early 2025, and is expected to reach more than 80,000 students and 7,000 faculty by the following year.

Additionally, UNSW is also exploring other uses of AI in the learning ecosystem to understand the added value it can bring. “We are also prototyping an orchestrator-style chatbot architecture based on multiple AI bots acting as personal concierges,” says McIntyre. “We have already begun a small pilot project later this year looking at using AI bots for roles such as administrative support for students, academic support in interpreting course information and lecture notes, and future student recruitment.”

The pilot of this chatbot technology will be assessed for its suitability as the primary interface for data analysis for student learning and support projects, and hopefully other university functions.

“Everything we're exploring in this project provides a great case study to help universities think about how to leverage the power of AI at scale,” McIntyre said. “This project has been a great way to bring people together to discover the potential of AI and help move the university forward.”

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *