Recommender Systems for Professional Development—Insights and Experiences from Practice

Abstract In the German market of continuing vocational education and training, millions of learners encounter countless continuing vocational education and training (VET) providers offering a wide range of courses. To help learners select suitable options, recommender systems are increasingly being used in digital continuing education searches. Recommender systems are software systems that guide users in a personalised way to interesting or useful items in a large range of possible options (Burke et al. in Aimag 32:13–18, 2011). On VET platforms, they can provide targeted recommendations for learning opportunities based, for example, on the skills and learning needs of learners (Drachsler et al. in Recommender systems handbook. Boston, MA, Springer US, 421–451, 2015; Reichow et al. in Recommender system in der beruflichen Weiterbildung. Grundlagen, Herausforderungen und Handlungsempfehlungen. Berlin, 2022). Most studies on recommender systems in education consider the higher education context (e.g. Deschênes in Int J Educ Technol Higher Educ 17(1) 2020)). VET, on the other hand, has been less researched (see, for example, Rivera et al., Proceedings of the International Conference on Information Technology & Systems (ICITS 2018). Springer International Publishing,, pp 937–947, 2018), leaving little systematic knowledge about the use of recommender systems in vocational education and training. The INVITE contest for digital innovations provided an opportunity to change this. In this BMBF-funded programme, more than 30 projects developed recommender systems for VET (BIBB, 2024; Reichow et al., Communications in Computer and Information Science. Springer, 2025). This article examines the implementation and potential of these recommender systems and addresses the following key questions: For what purpose and based on what technical methods were the recommender systems used in the INVITE projects? What was the pedagogical or didactic background? Based on this analysis and further literature research, the article identifies trends, areas for action and funding opportunities for future projects to further improve the fit between learners and digital educational content—in particular through the integration of generative AI—without neglecting critical aspects such as bias and data protection issues.