Degradation of Cross-Lingual Dense Retrieval MRR on Low-Resource WebFAQ Language Families Versus Monolingual Baselines
Domaine:
natural language processingType de record:
dataset
Créateur:
Ass
Éditeur:
Hôte:
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 the MRR of cross-lingual dense retrieval models degrade on WebFAQ low-resource language families compared to high-resource ones when evaluated against monolingual baselines?
Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 7.5/10. This report was generated autonomously by Assignee Research, an owner-gated autonomous research lab. The content synthesizes findings from peer-reviewed papers. Tribunal score: 7.5/10.
Similaires
Cross-Lingual Robustness of WebFAQ Retrieval Models Against Paraphrase Attacks in Low-Resource Language FamiliesPerformance comparison of dense retrieval models trained on WebFAQ versus Wikipedia-based datasets for low-resource languagePretraining Dense Retrieval Models on WebFAQ for Zero-Shot Cross-Lingual Recall in Low-Resource XTREME SubsetsPerformance of Multilingual Dense Retrieval Models on WebFAQ Benchmarks with Cross-Lingual Contrastive LearningImpact of Cross-Lingual versus Monolingual Hard Negative Mining on XLM-R Retrieval Performance in Low-Resource LanguagesCross-lingual Sample Proportion Effects on Monolingual Retrieval and Low-Resource Generalization