An Enhanced Span-based Decomposition Method for Few-Shot Sequence Labeling

Few-Shot Sequence Labeling (FSSL) is a canonical paradigm for the tagging models, e.g., named entity recognition and slot filling, to generalize on an emerging, resource-scarce domain. Recently, the metric-based meta-learning framework has been recognized as a promising approach for FSSL. However, most prior works assign a label to each token based on the token-level similarities, which ignores the integrality of named entities or slots. To this end, in this paper, we propose ESD, an Enhanced Span-based Decomposition method for FSSL. ESD formulates FSSL as a span-level matching problem between test query and supporting instances. Specifically, ESD decomposes the span matching problem into a series of span-level procedures, mainly including enhanced span representation, class prototype aggregation and span conflicts resolution. Extensive experiments show that ESD achieves the new state-of-the-art results on two popular FSSL benchmarks, FewNERD and SNIPS, and is proven to be more robust in the nested and noisy tagging scenarios. Our code is available at

PDF Abstract NAACL 2022 PDF NAACL 2022 Abstract


Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Few-shot NER Few-NERD (INTER) ESD 5 way 1~2 shot 59.29±1.25 # 3
5 way 5~10 shot 69.06±0.80 # 3
10 way 1~2 shot 52.16±0.79 # 4
10 way 5~10 shot 64.00±0.43 # 3
Few-shot NER Few-NERD (INTRA) ESD 5 way 1~2 shot 36.08±1.60 # 5
5 way 5~10 shot 52.14±1.50 # 4
10 way 1~2 shot 30.00±0.70 # 5
10 way 5~10 shot 42.15±2.60 # 5