基于因果干预的少样本命名实体识别（英文版）

英文标题：Causal Interventions-based Few-Shot Named Entity Recognition中文摘要：本文提出一种基于因果干预的少样本 NER 方法，通过背门调整和增量学习干预原型，避免了少样本选择偏差所带来的虚假相关性问题，并在不同基准测试中取得了新的最佳表现。英文摘要：Few-shot named entity recognition (NER) systems aims at recognizing newclasses of entities based on a few labeled samples. A significant challenge inthe few-shot regime is prone to overfitting than the tasks with abundantsamples. The heavy overfitting in few-shot learning is mainly led by spuriouscorrelation caused by the few samples selectio