Densifying Knowledge Hypergraphs with LLM-based review distillation for conversational consumer decision support
• Review-driven structural induction alleviates sparsity in conversational recommendation. • LLM-distilled review semantics are grounded as entities to densify hypergraph structures. • Item-centric hypergraph learning supports joint recommendation and response generation. • Two-stage training stabilizes structure learning under sparse conversational supervision.
Authors
- Francisco Javier Blanco‐Encomienda (ORCID: https://orcid.org/0000-0001-7449-3299)
- Yiyun Zhang (ORCID: https://orcid.org/0000-0002-1419-485X)
- Keyan Jin (ORCID: https://orcid.org/0009-0004-9783-8818)
Institutions
- Universidad de Granada (ES)
- Cathay Photonics (China) (CN)
Publication Details
- Journal
- Information Processing & Management
- Published
- 2026-05-22
- DOI
- https://doi.org/10.1016/j.ipm.2026.104919
- Primary Topic
- Sentiment Analysis and Opinion Mining
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Universidad de Granada