Interview with Fulbright scholar David Azcona - taking Insight to Arizona
Sharing Irish data insights with US universities - Insight's David Azcona is Ireland's latest Fulbright success
It’s one of the most prestigious scholarship programmes in the world, with alumni going on to win Nobel and Pulitzer prizes. The Fulbright Scholarship programme is a US talent outreach programme that sees 8000 scholars, artists and scientists each year take their skills to a new jurisdiction to share and exchange knowledge. The relationship between Ireland and Fulbright is sixty years old this year, and hundreds of students in Ireland have taken their skills to universities abroad; to share their expertise, enrich their studies and build lasting, positive global networks based on shared values.
The latest awardee is Insight Centre for Data Analytics researcher David Azcona. David has taken a year out of his Dublin City University PhD programme to apply his research in Arizona State University. David is using data to support university students and improve academic outcomes. His work caught the eye of the Fulbright selection committee due to measurable success in the DCU Computer Science School, with increased retention rates and improved exam results for students at the North Dublin campus.
“I’m interested in how the data that we use for commercial activities such as hotel and film recommendations could be applied to the field of education,” says Azcona, who is from Pamplona in Spain and has been working and studying in Ireland for five years.
“I developed this interest during a two year masters programme in data analytics at IBM, and decided to explore the area further with the Insight Centre at DCU. Over the least two years I have been working with students of computer science, using data to personalise their learning, recommend suitable study materials and offer solutions to failed submissions. We have been able to demonstrate improvements for students in DCU using machine learning and artificial intelligence.”
With the development of student intranet systems for the delivery of course content, academic activity is increasingly conducted online. This creates opportunities for the use of machine learning to capitalise on information captured when students interact with course content. It also provides the means to personalise recommendations for individuals. In much the same way that users of TV streaming services are guided in their viewing options according to previously viewed materials, knowledge of student interactions with materials can generate individual guides to optimal study plans.
“I use data mining techniques to run predictions on student outcomes and generate prescribed recommendations. I email students at the end of each week with the results. I’ve sent out thousands of notifications since my research began,” David explains. “I also send reports to lecturers each week and I have a dashboard that lecturers can use to get an overview of their class performance, and predicted outcomes.”
The work also supports peer learning – an important element of any successful programme. “Those students that don’t do well on a programming assignment can learning from those who got a high ranking. My work helps to link students to best practice within their programme.”
Computer science students in Ireland are vulnerable to non-completion, with The Irish Times reporting drop out rates of up to 70 per cent on some courses. Overall, about one in six, or just over 6,000, students did not progress to second year in 2015/16. Computer science, construction and business courses recorded some of the highest levels of non-progression.
David’s work has the potential to support vulnerable students in a field of study that is very important to the Irish economy. His Fulbright success points to a similar priority in the US. “The artificial intelligence approach and methodologies I am using can be applied in any academic setting. In Arizona State University I am working with much larger datasets, allowing me to create better data models. Privacy rules in the US mean that I will potentially have more freedom around data use and sharing and can develop more complicated models such as adaptive assignments. I will bring this learning back with me to Ireland.”
This is the beauty of the Fulbright model – the dissemination and development of exciting research across jurisdictions. It’s also a chance for students to see the world. He’s very grateful for the opportunity and has had a very productive year so far.
‘I’m researching learner behaviour in Computer Science studies and the differences between the approaches in Ireland and America, and how students can work and collaborate with a university the size of ASU,” David explains. ‘I am sponsored by John Rome, Deputy CIO at ASU, and Arthur Blakemore, Vice-President for Student Success at ASU Provost Office.
‘I’m working in conjunction with the ASU's University Technology Office (UTO), ASU Online EdPlus, and ASU’s School of Computing, Informatics, and Decision Systems Engineering. I had already been working on the relationship between learners and Computer Science curriculum at DCU using artificial intelligence and data-driven approaches, under Processor Alan Smeaton. ASU combines a large student body with innovative approaches in education, a perfect symbiosis for my research. I am very excited to be here for the year!’
At the moment David is examining how DCU and ASU approaches to guiding students differ, recommending study paths and working with professors to provide a solid framework for computer science studies. The collaboration with ASU's UTO, their university tech shop, is allowing him to gather a rich student digital footprint at the university and potentially input the insights extracted into the decisions being made for students in DCU and ASU.
‘In March, I presented a joint workshop paper in the 8th International Learning Analytics & Knowledge Conference (LAK 2018) in Sydney (Australia) with Sharon Hsiao, Assistant Professor at ASU's School of Computing, and that extended paper has been accepted to International Journal of Artificial Intelligence in Education (ijAIED). Prof Hsiao runs the Computing Systems & Informatics research lab where I have been collaborating on Educational Data Mining projects,’ David reveals.
His scholarship has given him lots of travel opportunities, he says. ‘In the US, I attended Amazon's ReInvent 2017 conference in Las Vegas and the Open Data Science Conference in San Francisco. I will be travelling to the Fulbright Enrichment Seminar in Philadelphia and a Higher Education conference in Oregon in April. It has been an amazing experience so far thanks to my sponsor John Rome here, the Fulbright Program and everybody I met here at ASU.
“I’m also playing soccer with ASU, as I did while t DCU. The 45 degree heat is sometimes a challenge but I come from Spain so it’s not completely foreign to me!” he says.
Will he return to Ireland after his year in Arizona? “Certainly – I believe working with the large ASU student body is going to provide me with very valuable observations that I can bring to my work in the Insight Centre at DCU. There is also scope to extend the programme beyond the computer science department into other subject areas and I would like support that development. I lived in Ireland for five years and the opportunities for data research there are amazing. I’m proof of that.”
David Azcona is co-organising a workshop on Educational Data Mining in Computer Science Education (CSEDM) in conjunction with EDM 2018 at The University at Buffalo, New York, United States, July 15th, 2018. Workshop details and submission format are at https://sites.google.com/asu.