Fine-Tuning Dense Retrieval on WebFAQ Non-English Subsets for Zero-Shot Cross-Lingual MRR@10 in Low-Resource Languages

We present WebFAQ, a large-scale collection of open-domain question answering datasets derived from FAQ-style schema.org annotations. In total, the data collection consists of 96 million natural question-answer (QA) pairs across 75 languages, including 47 million (49\%) non-English samples. WebFAQ further serves as the foundation for 20 monolingual retrieval benchmarks with a total size of 11.2 million QA pairs (5.9 million non-English). These datasets are carefully curated through refined filtering and near-duplicate detection, yielding high-quality resources for training and evaluating multil

Research goal: How does fine-tuning dense retrieval models on WebFAQ's non-English subsets impact zero-shot cross-lingual transfer MRR@10 performance on low-resource language benchmarks?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 8.6/10.