From Data Enrichment to Responsible Recommendations: Addressing Heterogeneous Information Insufficiency in Recommender Systems
Kim Thanh Vu
This thesis addresses heterogeneous information insufficiency in recommender systems, the problem that collaborative signals alone are too sparse to support accurate and responsible recommendation. It introduces a unifying framework spanning contextual, semantic, multimodal, reliability, and fairness dimensions, and de...
- Queensland University of Technology
- DOI: 10.5204/thesis.eprints.267923