Research map / Large language models / AI and society
TrueLearn: A Family of Bayesian Algorithms to Match Lifelong Learners to Open Educational Resources
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add a novelty margin so engagement reflects both learner knowledge and resource novelty (ablation) | TrueLearn · 20 learners-5 topics | F1: .790 -> .821 | Helped |
| add a novelty margin so engagement reflects both learner knowledge and resource novelty (ablation) | TrueLearn · All learners-5 topics | F1: .626 -> .677 | Helped |
| use cosine similarity as fixed resource depth instead of inferring depth (ablation) | TrueLearn · All learners-5 topics | F1: .418 -> .626 | Helped |
| exclude negative engagement observations when updating learner knowledge (ablation) | TrueLearn fixed-depth · 20 learners-5 topics | F1: .790 -> .813 | Helped |
| exclude negative engagement observations when updating learner knowledge (ablation) | TrueLearn fixed-depth · 20 learners-10 topics | F1: .761 -> .783 | Helped |
| exclude negative engagement observations when updating learner knowledge (ablation) | TrueLearn fixed-depth · All learners-5 topics | F1: .573 -> .626 | Helped |
| increase the resource representation from 5 to 10 topics (ablation) | TrueLearn fixed-depth · 20 learners | F1: .783 -> .761 | Hurt |
| increase the resource representation from 5 to 10 topics (ablation) | TrueLearn Novelty · 20 learners | F1: .821 -> .784 | Hurt |
| add a Gaussian drift term to learner skills (ablation) | TrueLearn and other models · VideoLectures.net datasets | F1 | No clear effect |
| use multiple Wikipedia-topic skills rather than one skill per whole video (ablation) | TrueLearn Novelty · 20 most active users, first 100 events | F1: 0.855 vs 0.853 | No clear effect |
| use multiple Wikipedia-topic skills rather than one skill per whole video (ablation) | TrueLearn Novelty · 20 most active users, last 100 events | F1: 0.783 -> 0.796 | Helped |
| use multiple Wikipedia-topic skills rather than one skill per fragment (ablation) | Vanilla TrueSkill Video · All learners-5 topics | F1: .480 -> .679 | Helped |
Closest papers by meaning
- Multi-type MOOCs Recommendation: Leveraging Deep Multi-Relational Representation and Hierarchical Reasoning (AAAI 2025 · similarity 0.88)
- A Toolbox for Modelling Engagement with Educational Videos (AAAI 2024 · similarity 0.87)
- Personalized Exercise Recommendation with Semantically-Grounded Knowledge Tracing (NeurIPS 2025 · similarity 0.87)
- GenAL: Generative Agent for Adaptive Learning (AAAI 2025 · similarity 0.86)
- Predictive, scalable and interpretable knowledge tracing on structured domains (ICLR 2024 · similarity 0.84)
- Influential Exemplar Replay for Incremental Learning in Recommender Systems (AAAI 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.