{"ID":2837602,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.19176","arxiv_id":"2511.19176","title":"From Raw Features to Effective Embeddings: A Three-Stage Approach for Multimodal Recipe Recommendation","abstract":"Recipe recommendation has become an essential task in web-based food platforms. A central challenge is effectively leveraging rich multimodal features beyond user-recipe interactions. Our analysis shows that even simple uses of multimodal signals yield competitive performance, suggesting that systematic enhancement of these signals is highly promising. We propose TESMR, a 3-stage framework for recipe recommendation that progressively refines raw multimodal features into effective embeddings through: (1) content-based enhancement using foundation models with multimodal comprehension, (2) relation-based enhancement via message propagation over user-recipe interactions, and (3) learning-based enhancement through contrastive learning with learnable embeddings. Experiments on two real-world datasets show that TESMR outperforms existing methods, achieving 7-15% higher Recall@10.","short_abstract":"Recipe recommendation has become an essential task in web-based food platforms. A central challenge is effectively leveraging rich multimodal features beyond user-recipe interactions. Our analysis shows that even simple uses of multimodal signals yield competitive performance, suggesting that systematic enhancement of...","url_abs":"https://arxiv.org/abs/2511.19176","url_pdf":"https://arxiv.org/pdf/2511.19176v3","authors":"[\"Jeeho Shin\",\"Kyungho Kim\",\"Kijung Shin\"]","published":"2025-11-24T14:37:22Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.IR\"]","methods":"[]","has_code":false}