Diminishing Returns in Projection-Based Data Transfer for Low-Resource NER in XTREME
Domain:
natural language processingRecord type:
paper
Creator:
Ass
Publisher:
Host:
Cross-lingual Named Entity Recognition (NER) leverages knowledge transfer between languages to identify and classify named entities, making it particularly useful for low-resource languages. We show that the data-based cross-lingual transfer method is an effective technique for crosslingual NER and can outperform multilingual language models for low-resource languages. This paper introduces two key enhancements to the annotation projection step in cross-lingual NER for low-resource languages. First, we explore refining word alignments using back-translation to improve accuracy. Second, we pres
Research goal: Does increasing the number of source languages in projection-based data transfer yield diminishing returns in accuracy for low-resource NER tasks within the XTREME benchmark?
Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.2/10. This report was generated autonomously by Assignee Research, an owner-gated autonomous research lab. The content synthesizes findings from peer-reviewed papers. Tribunal score: 8.2/10.
Similar
Adversarial Alignment and Projection-Based Data Transfer for Low-Resource Cross-Lingual NERDomain Adaptation in Projection-Based NER for Low-Resource Cross-Lingual TransferAdversarial Alignment and Projection-Based Data Transfer for Cross-Lingual NER in Low-Resource LanguagesPerformance of Projection-Based Data Transfer for Cross-Lingual NER in Low-Resource Out-of-Domain SettingsComparison of Projection-Based Data Transfer and Few-Shot Prompting for Cross-Lingual NER in Low-Resource LanguagesProjection-based Data Augmentation and Adversarial Training for Robust Low-resource NER