Invited talk at the Teaching Machine Learning Workshop (satellite event) at ECML 2022

Septembar 13, 2022

https://teaching-ml.github.io/2022/

Abstract

Subjects like machine learning and data science are changing very fast. It’s thus a challenge for any particular department or college to teach high-quality and up-to-date courses on these topics. By looking to adopt ethos and processes from open source software and open science, we can enhance quality and outcomes in teaching new subjects. Ethos means the practices and values that characterize open source communities. The open education movement starting in the 1990s was inspired by open source software; its most visible efforts have been open courseware (OCW) and open educational resources (OER). But key features were missed: the open development model, community building, and networked collaboration. Teaching in the Open means looking to form collaborations in the development of curricula and content, sharing learning objects under permissive licenses, thinking of reuse from the beginning, participating in peer review of content and learning objects, and accepting community contributions. In this vein, we founded The Journal of Open Source Education (https://jose.theoj.org), publishing papers on both software for educational purposes, and learning modules, particularly on computing-based courses. The contributions of authors, editors, and reviewers show that communities of teacher-scholars are forming, growing, and having impact.