Masterthesis

Data-Driven Embedding of Educational Resources in a Vector Space with Interpretable Dimensions for Explainable Recommendation

Repos and Pages

• High level introduction on this page: here
• Code: https://github.com/cstenkamp/derive_conceptualspaces
• Text-Repo: https://github.com/cstenkamp/MastersThesisText
• Text as PDF: interactive, printable
• Analysis-Notebooks: https://github.com/cstenkamp/MAAnalysisNotebooks
• Referenced Plots: open in nbviewer or binder
• Snakemake SGE-IKW-Grid-Profile: https://github.com/cstenkamp/Snakemake-IKW-SGE-Profile

Abstract

The number of available educational resources keeps rising just as the liberty for students to create a custom curriculum, making recommendation in this sector ever more critical. There is the need for a methodology to structure these resources unsupervisedly to allow for explainable recommendations, where users can explicitly demand features such as the degree how mathematical a course should be. Conceptual Spaces allow to embed resources into a vector space whose quality dimensions correspond to salient semantic features, but generating them reliably is an open question. This thesis replicates a method to create these spaces data-driven from their descriptions and applies it to a given dataset of that domain to explore its applicability. Our preliminary results show that human categories such as the faculty a course belongs to are among the detected salient directions. Further qualitative analysis also indicates that the embeddings capture essential features that can be used in interpretable classification and explainable recommendation. Additionally, a scalable and modular architecture for the algorithm is created. It is shown that it can be applied to diverse variations of dataset and algorithm components. A workflow is implemented and described to run the algorithm efficiently on compute clusters. The full implementation is open-sourced to make future research on related problems more accessible.