Improving Korean-English Cross-Lingual Retrieval: A Data-Centric Study of Language Composition and Model Merging

2025/07/11 by Jang, Youngjoon, Son, Junyoung, Lee, Taemin +2
#FOS: Computer and information sciences #Information Retrieval (cs.IR)

paper · doi:10.48550/arxiv.2507.08480

Abstract

With the increasing utilization of multilingual text information, Cross-Lingual Information Retrieval (CLIR) has become a crucial research area. However, the impact of training data composition on both CLIR and Mono-Lingual Information Retrieval (IR) performance remains under-explored. To systematically investigate this data-centric aspect, we construct linguistically parallel Korean-English datasets and train retrieval models with various language combinations. Our experiments reveal that the language composition of training data significantly influences IR performance, exhibiting important inter-lingual correlations: CLIR performance improves with specific language pairs, while Mono-Lingual IR performance declines. Our work demonstrates that Model Merging can effectively mitigate this trade-off, achieving strong CLIR results while preserving Mono-Lingual IR capabilities. Our findings underscore the effects of linguistic configuration of training data on both CLIR and Mono-Lingual IR, and present Model Merging as a viable strategy to optimize performance across these tasks.

Citations

• Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models
• MMTEB: Massive Multilingual Text Embedding Benchmark
• Agentic Retrieval-Augmented Generation: A Survey on Agentic RAG
• mGTE: Generalized Long-Context Text Representation and Reranking Models for Multilingual Text Retrieval
• NV-Retriever: Improving text embedding models with effective hard-negative mining
• From Matching to Generation: A Survey on Generative Information Retrieval
• From Matching to Generation: A Survey on Generative Information Retrieval
• M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation
• Large Language Models for Information Retrieval: A Survey
• English Contrastive Learning Can Learn Universal Cross-lingual Sentence Embeddings
• Neural Approaches to Multilingual Information Retrieval
• Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time
• Mind the Gap: Cross-Lingual Information Retrieval with Hierarchical Knowledge Enhancement
• On Cross-Lingual Retrieval with Multilingual Text Encoders
• Merging Models with Fisher-Weighted Averaging
• SPLADE: Sparse Lexical and Expansion Model for First Stage Ranking
• Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup
• Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategies
• Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval
• Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
• ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT
• Dense Passage Retrieval for Open-Domain Question Answering
• Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
• A Theoretical Analysis of NDCG Type Ranking Measures

Related

• What Drives Cross-lingual Ranking? Retrieval Approaches with Multilingual Language Models
• Bridging Language Gaps: Advances in Cross-Lingual Information Retrieval with Multilingual LLMs
• On Cross-Lingual Retrieval with Multilingual Text Encoders
• Language Agnostic Multilingual Information Retrieval with Contrastive Learning
• Simple Yet Effective Neural Ranking and Reranking Baselines for Cross-Lingual Information Retrieval