Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.01566 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918143595642880 |
|---|---|
| author | Wang, Yujing Chen, Yiren Li, Huoran Xu, Chunxu Luo, Yuchong Mao, Xianghui Li, Cong Du, Lun Ma, Chunyang Jiang, Qiqi Wang, Yin Gao, Fan Mo, Wenting Wen, Pei Kumar, Shantanu Park, Taejin Song, Yiwei Rajaram, Vijay Cheng, Tao Durgia, Sonu Kolari, Pranam |
| author_facet | Wang, Yujing Chen, Yiren Li, Huoran Xu, Chunxu Luo, Yuchong Mao, Xianghui Li, Cong Du, Lun Ma, Chunyang Jiang, Qiqi Wang, Yin Gao, Fan Mo, Wenting Wen, Pei Kumar, Shantanu Park, Taejin Song, Yiwei Rajaram, Vijay Cheng, Tao Durgia, Sonu Kolari, Pranam |
| contents | As global e-commerce platforms continue to expand, companies are entering new markets where they encounter cold-start challenges due to limited human labels and user behaviors. In this paper, we share our experiences in Coupang to provide a competitive cold-start performance of relevance matching for emerging e-commerce markets. Specifically, we present a Cold-Start Relevance Matching (CSRM) framework, utilizing a multilingual Large Language Model (LLM) to address three challenges: (1) activating cross-lingual transfer learning abilities of LLMs through machine translation tasks; (2) enhancing query understanding and incorporating e-commerce knowledge by retrieval-based query augmentation; (3) mitigating the impact of training label errors through a multi-round self-distillation training strategy. Our experiments demonstrate the effectiveness of CSRM-LLM and the proposed techniques, resulting in successful real-world deployment and significant online gains, with a 45.8% reduction in defect ratio and a 0.866% uplift in session purchase rate. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_01566 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CSRM-LLM: Embracing Multilingual LLMs for Cold-Start Relevance Matching in Emerging E-commerce Markets Wang, Yujing Chen, Yiren Li, Huoran Xu, Chunxu Luo, Yuchong Mao, Xianghui Li, Cong Du, Lun Ma, Chunyang Jiang, Qiqi Wang, Yin Gao, Fan Mo, Wenting Wen, Pei Kumar, Shantanu Park, Taejin Song, Yiwei Rajaram, Vijay Cheng, Tao Durgia, Sonu Kolari, Pranam Information Retrieval Computation and Language As global e-commerce platforms continue to expand, companies are entering new markets where they encounter cold-start challenges due to limited human labels and user behaviors. In this paper, we share our experiences in Coupang to provide a competitive cold-start performance of relevance matching for emerging e-commerce markets. Specifically, we present a Cold-Start Relevance Matching (CSRM) framework, utilizing a multilingual Large Language Model (LLM) to address three challenges: (1) activating cross-lingual transfer learning abilities of LLMs through machine translation tasks; (2) enhancing query understanding and incorporating e-commerce knowledge by retrieval-based query augmentation; (3) mitigating the impact of training label errors through a multi-round self-distillation training strategy. Our experiments demonstrate the effectiveness of CSRM-LLM and the proposed techniques, resulting in successful real-world deployment and significant online gains, with a 45.8% reduction in defect ratio and a 0.866% uplift in session purchase rate. |
| title | CSRM-LLM: Embracing Multilingual LLMs for Cold-Start Relevance Matching in Emerging E-commerce Markets |
| topic | Information Retrieval Computation and Language |
| url | https://arxiv.org/abs/2509.01566 |