This article was featured on University World News webpage on November 27, 2024.

Southeast Asia’s first data capacity accelerator, which will be used to train students and data professionals to use data for social impact, was inaugurated in Singapore this month by data.org, a non-governmental organisation, in partnership with the Association of Pacific Rim Universities (APRU) and the National University of Singapore (NUS).

The initiative will produce a cohort of socially responsible data professionals and a social impact training model for universities that can be scaled up in about two years across other Southeast Asian countries, APRU, stated.

“Digital transformation, AI and data all have a role to play in shaping society and driving economies towards financial health and resilience,” said Shamina Singh, founder and president of the Mastercard Center for Inclusive Growth, which is funding the initiative.

She underlined the need to “reach the next generation of data practitioners, so they can harness the power of data and AI to support inclusive economic growth in the region”.

Jon Huntsman, a former US ambassador to China and Singapore and vice-chairman (strategic growth) at Mastercard, which oversees the centre, said that the accelerator launch was “a remarkable step in the region’s journey as a leader in using data and AI for social good. This initiative will prepare a new generation of data and AI practitioners in the civil service, among others, to address social challenges that benefit local communities”.

Professor Huang Ke-Wei, executive director of NUS’ Asian Institute of Digital Finance (AIDF), at the 22 November launch described the accelerator as “a programme designed to uplift under-represented communities across Asia by equipping them with the skills and tools needed to thrive in an increasingly data-driven world. This mission has never been more important … as AI continues to evolve at a very rapid pace”.

Tapping into the power of higher education

APRU chief executive Thomas Schneider said: “This partnership is about tapping into the power of higher education to ensure that our workforce and our communities are not left behind.”

With its membership of over 60 universities in Asia and the Pacific, North America and Latin America, APRU “would like to create open-source resources in data science, where people can be trained at a massive scale. And we would also want to tie in many of our [APRU’s] huge programme hubs in global health, in sustainability and in food security, so students can learn practice-based solutions”, he noted.

“Data science for social impact has the potential for significant societal benefits in areas such as economic mobility, gender equity, and even public health and climate, so we are eager to see how the data practitioners and social impact organisations involved will address this challenge in a way that serves the public good in the Asia Pacific and beyond,” Schneider added.

Data for impact

Data.org, a platform for partnerships between philanthropy, technology, academia, and social impact organisations, specifically focuses on data science for social impact. It has already established hubs at universities in Africa, India, Latin America, and the United States.

Executive director of data.org Danil Mikhailov said that the aim is to apply research and academic expertise to enable social impact organisations “to unlock the power of data to meet their missions”.

He told University World News: “Our focus is to democratise access to these skills but also make sure that data science is taught in a new way – with a focus on responsible data and responsible AI. And for it to be responsible, the skills you need are not just technical skills like coding. You need to understand how to work with vulnerable communities.”

He noted that it required skills from the humanities and social sciences. For example: “If you want to work in health, you need to understand the basics of how health data is different from climate data or financial inclusion data,” he said.

“We want people to be interdisciplinary scholars who approach data in what we call a socio-technical way – society and technology, not just technology,” he explained.

Data.org has a goal to train 1 million data scientists, data analysts, data engineers around the world by 2032, particularly focusing on areas that are currently excluded. But Mikhailov said this could not be achieved with partnerships with individual universities if it is to have a global impact, hence the need for a ‘data capacity accelerator’ scheme.

A different type of university course

AIDF in Singapore, which involves collaboration between industry, academia and government policymakers, has devised a course on inclusive finance and will develop many more under the new initiative. The ‘data capacity accelerator’ is intended to address financial inequality in society and create meaningful social impact, AIDF’s Huang said.

As part of the partnership, AIDF will co-create open-source certification courses in AI and data analytics, and through the initiative reach learners in other countries “to ensure that data and AI are harnessed equitably to empower lives and strengthen underprivileged communities”, he noted.

He said at the launch: “We try to empower communities to reduce inequalities and also drive sustainable growth through digital finance and AI innovations.”

Mikhailov stated that the students “have been taught a more responsible approach to data science with a social impact lens, and that embeds a new culture, a little seed of culture that will germinate in those institutions”.

He noted: “Then we pair institutions with social impact organisations. So when those students come out, they get an experience of what it’s like on the other side, where the resources are much tighter, where the problems are much bigger. And that, again, informs them and shapes them into the well-rounded, interdisciplinary data scientists of the future that we all need.”

Course design

Referring to the need for exercises, case studies and data sets used by students to be derived from “real world problems”, Mikhailov said that it was important to ensure that course materials co-designed with universities had a strong applied element.

All too often, data scientists being trained on AI models for the first time are given inappropriate data sets, Mikhailov noted. These might include data sets from retail – something that’s easily accessible, “but you’re actually trying to solve a problem in health or climate, so the data is very different and therefore you’re not learning the right skills and approaches”, he explained.

Many students at data capacity accelerator initiatives elsewhere in the world work with local social impact organisations near the university. “We provide fellowship opportunities for students, fully paid by us, to go in and work in those organisations, let’s say, for a year, applying their skills,” Mikhailov said.

“That’s good for students because they get a practical test of what they’ve learnt, and they learn so much more by doing, rather than by rote, and it’s very good for the organisations because often they’re priced out in the marketplace, unable to afford data scientists,” he added.

Some of the 20 universities linked to data.org around the world created a new undergraduate degree; others, such as NUS, embed specific modules into existing courses, while others co-design modules with data.org.

“The materials they [universities] produce, case studies and various teaching assets, are then used within our network and accessed by many more players who use them in their own courses.”

Spreading in the region

Christina Schönleber, APRU’s chief strategy officer, said that APRU would build more co-designed courses and learning materials to roll out to the region.

“Our member universities can leverage their deep understanding of ocal contexts, communities and challenges to ensure the training programmes and resources are culturally and contextually relevant,” Schönleber told University World News.

“Then, over the course of next year, develop and also run open-source courses and related materials. We need to develop these and make them available for students, academics and organisations who do not yet understand it [the application of data with impact].

“We also need to make sure the courses can be taken up by another university and they can integrate these, or elements of these, within their curricula,” she noted.

APRU also intends to use some of the courses already developed under data.org, including some on the basics of data science.

APRU’s plan is to launch a social impact data accelerator innovation platform, a practice-based learning platform “to get computer and data science students to work with social impact organisations across the Asia Pacific region, to help students understand how the use and application of data can make a real difference”, she explained.

Many academics from the APRU network and beyond already work together on APRU programmes. “We bring academics together that work within the communities and work on the ground,” she said, and some will be brought in to work with computer and data scientists for the new initiative.

“That’s the biggest challenge: to build that interdisciplinary community across the region to start with,” she said.

“They will hopefully continue to collaborate beyond this initiative. Part of the legacy will be open-source course materials and assets, and we will work with data.org, and our other partners to share them widely.”