OSDG and University of Hong Kong: OSDG Community Dataset used as a training dataset
March 30, 2022
The OSDG team has recently had an exciting opportunity to assist members of the University of Hong Kong’s Technology-Enriched Learning Initiative (TELI) and the University’s Common Core Office.
TELI consists of a “team of e-learning technologists, instructional designers, multimedia professionals, software engineers, game developers, researchers and collaboration associates under the Teaching and Learning infrastructure of the University of Hong Kong” (source). The team is actively using technology as a tool to scale up teaching and learning practices within the university and beyond using Massive Open Online Courses (MOOCs), Small Private Online Courses (SPOCs), and other e-learning technologies, activities, and projects.
The TELI team has recently set out to evaluate SDG-related formal learning activities in the university’s common core curriculum, and to contribute to currently limited research on how universities can holistically promote the Sustainable Development Goals (SDGs) through their curriculum.
“The Common Core at HKU is committed to responding purposefully to the wide-ranging challenges identified by the United Nations through the Sustainable Development Goals. In both curricular and co-curricular ways, we work with students, colleagues, and partners to deepen our engagement with the SDGs”, – says Professor Gray Kochhar-Lindgren, Director of the University of Hong Kong’s Common Core Office.
The TELI research was guided by the following questions:
1) How can common core courses be classified to the SDGs by utilising machine learning algorithms?
2) How have the SDGs been taught in the university common core courses so far, assessed from a course, curriculum, and theme cluster perspective?
3) What is the difference between SDG classifications conducted by a human vs a machine?
The researchers classified 166 common core courses with SDGs using both human and machine learning algorithms, while making use of the OSDG Community Dataset (OSDG-CD) as a training dataset. The OSDG-CD contains tens of thousands of texts, validated by the OSDG Community platform citizen scientists and volunteers with respect to SDGs and is publicly available on the Zenodo repository with quarterly updates.
Methodology of classifying courses to SDGs
The research team – Chi-Un Lei, Xinyi Liang, Chun Yin Cham, Xue Qian, and Xiaodong Hu –recently published a poster paper at the International Conference on Learning Analytics and Knowledge 2022, entitled “Assessing the Integration of United Nations Sustainable Development Goals in a University General Education Curriculum”, detailing the methodology and initial results. We invite you to access the paper and a short video presentation to find out more.
Contact us for similar use cases
The case of TELI is a great example of how SDGs can be assigned via machine learning, and used to obtain valid insights into a university curriculum. Here at OSDG, we are eager to find new applications of our methodology via our classification tool, API, Community Dataset or overall methodology. If you are interested in large-scale classification of SDGs, please get in touch with our team.
