Research map / Vision-language and multimodal / Cross-modal learning
Joint Representation Learning for Multi-Modal Transportation Recommendation
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 user and OD relevance regularization to the base embedding model (ablation) | Trans2Vec · BEIJING | NDCG@5: 0.876 -> 0.893 | Helped |
| add user and OD relevance regularization to the base embedding model (ablation) | Trans2Vec · BEIJING | REC: 0.736 -> 0.770 | Helped |
| add user and OD relevance regularization to the base embedding model (ablation) | Trans2Vec · SHANGHAI | F1: 0.718 -> 0.719 | Helped |
| use fixed anchor embeddings for modes with joint learning instead of PTE/Metapath2Vec embeddings | BTrans2Vec · BEIJING | NDCG@5: 0.770 -> 0.876 (vs PTE) | Helped |
| Trans2Vec instead of Metapath2Vec on precision | Trans2Vec · BEIJING | PREC: 0.718 -> 0.700 | Hurt |
| increase number of relevance neighbors K from 0 to 1 (ablation) | Trans2Vec · BEIJING | all metrics | Helped |
| increase K beyond 5-10 (up to 50) (ablation) | Trans2Vec · BEIJING | PREC | Hurt |
| add user relevance (beta1 0.1-0.3) (ablation) | Trans2Vec · BEIJING | all metrics | Helped |
| add OD relevance (beta2 0.3-1.0) (ablation) | Trans2Vec · BEIJING | all metrics | Helped |
| vary embedding dimension d from 16 to 256 (ablation) | Trans2Vec · BEIJING | all metrics | No clear effect |
| increase d from 16 to 32 (ablation) | Trans2Vec · BEIJING | REC vs PREC | Mixed |
Closest papers by meaning
- Learning Accurate and Bidirectional Transformation via Dynamic Embedding Transportation for Cross-Domain Recommendation (AAAI 2024 · similarity 0.84)
- LGMRec: Local and Global Graph Learning for Multimodal Recommendation (AAAI 2024 · similarity 0.83)
- Beyond Graph Convolution: Multimodal Recommendation with Topology-aware MLPs (AAAI 2025 · similarity 0.83)
- SPOT-Trip: Dual-Preference Driven Out-of-Town Trip Recommendation (NeurIPS 2025 · similarity 0.83)
- Semantic Routing via Autoregressive Modeling (NeurIPS 2024 · similarity 0.82)
- STAIR: Manipulating Collaborative and Multimodal Information for E-Commerce Recommendation (AAAI 2025 · similarity 0.82)
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.