Search papers, labs, and topics across Lattice.
This paper introduces a disentangled contrastive learning (DCL) approach for multilingual dense retrieval, addressing the challenge of transferring retrieval capabilities to low-resource languages where annotated data is limited. By separating semantic and linguistic features into distinct subspaces, the method minimizes interference during semantic matching, allowing for more effective retrieval across languages. Experimental results on mMARCO and MIRACL datasets reveal that DCL significantly outperforms existing baselines, showcasing its robustness and generalization potential in zero-shot scenarios.
Separating semantic and linguistic features can drastically enhance multilingual retrieval performance, especially in low-resource settings.
Multilingual dense retrieval aims to handle queries and documents across different languages based on a unified retriever model. The challenge lies in enabling robust retrieval transfer to low-resource languages where annotated retrieval data is often scarce. Although previous studies transfer high-resource supervision to low-resource languages in multilingual semantic representation learning, the shared representation often entangles semantic and linguistic features, which may interfere with optimizing semantic relevance for retrieval. Different from existing methods that focus on learning language-agnostic semantic features under such entanglement, we propose a disentangled contrastive learning~(DCL) method for multilingual dense retrieval by separating multilingual representations into semantic and linguistic subspaces. Specifically, we design disentangled optimization objectives based on hierarchical semantic alignment and language debiasing contrastive learning. By aligning retrieval-relevant semantics across languages at both sentence and token levels while capturing language-specific variations in the linguistic subspace, these objectives reduce language-induced interference in semantic matching. We jointly optimize them with the retrieval objective to facilitate stable zero-shot transfer from English supervision to multilingual dense retrieval. Extensive experiments on mMARCO and MIRACL show that our method consistently outperforms several strong baselines, demonstrating its effectiveness and generalization ability.
