The Campus Analytics project has its roots in the EPFL’s MOOCs initiative. In 2012, the Center for Digital Education (CEDE) was created to serve professors who wish to create online courses, and to make EPFL a world leader in this emerging area. As a result, our MOOCs amassed more than 2 million registrations.
This massive number of students, and log files produced by their usage of the platforms, gave us the first glimpse into educational big data. The MOOCs dataset allowed us to understand, through the use of data science and machine learning techniques, the effectiveness of different teaching methods and learning strategies (see publications).
It didn’t take long until we asked ourselves if this kind of analysis could be applied to our own students at EPFL, as an effort to improve their learning and the quality of our teaching. We came up with questions such as “Can we structure our course catalog in an optimal way, based on actual student performance?”, or “Can we recommend courses to our students based on their desired future career?”.
By unifying data sources at EPFL, we can perform this global analysis and extend it to other aspects of the university’s mission, such as scientific research and entrepreneurship. For instance, can we recommend potential collaborations to our scientists so that EPFL collectively increase its research output? Can we predict which research area or educational program is obtaining more investment for the creation of startups? (See example projects).
These are interesting research questions that we are trying to answer, both by creating a unified database of EPFL academic data and building partnerships with our own research labs, where people have the knowhow to apply advanced data science, machine learning, and deep learning algorithms to obtain insights from this data. These insights will be of great value to the EPFL administration, allowing us to make strategic decisions based on actual data.
If you’re a Ph.D. student, Postdoc, or Professor in a research lab at EPFL, and you wish to discuss a potential collaboration, please feel free to contact us.