Parallel Instance Query Network for Named Entity Recognition

Named entity recognition (NER) is a fundamental task in natural language processing. Recent works treat named entity recognition as a reading comprehension task, constructing type-specific queries manually to extract entities. This paradigm suffers from three issues. First, type-specific queries can only extract one type of entities per inference, which is inefficient. Second, the extraction for different types of entities is isolated, ignoring the dependencies between them. Third, query construction relies on external knowledge and is difficult to apply to realistic scenarios with hundreds of entity types. To deal with them, we propose Parallel Instance Query Network (PIQN), which sets up global and learnable instance queries to extract entities from a sentence in a parallel manner. Each instance query predicts one entity, and by feeding all instance queries simultaneously, we can query all entities in parallel. Instead of being constructed from external knowledge, instance queries can learn their different query semantics during training. For training the model, we treat label assignment as a one-to-many Linear Assignment Problem (LAP) and dynamically assign gold entities to instance queries with minimal assignment cost. Experiments on both nested and flat NER datasets demonstrate that our proposed method outperforms previous state-of-the-art models.

PDF Abstract ACL 2022 PDF ACL 2022 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Nested Named Entity Recognition ACE 2004 PIQN F1 88.14 # 6
Named Entity Recognition (NER) ACE 2004 PIQN F1 88.14 # 2
Multi-Task Supervision n # 1
Nested Named Entity Recognition ACE 2005 PIQN F1 87.42 # 3
Named Entity Recognition (NER) ACE 2005 PIQN F1 87.42 # 4
Named Entity Recognition (NER) CoNLL 2003 (English) PIQN F1 92.87 # 32
Named Entity Recognition (NER) Few-NERD (SUP) Locate and Label Precision 64.69 # 6
Recall 70.87 # 2
F1-Measure 67.64 # 5
Named Entity Recognition (NER) Few-NERD (SUP) Sequence-to-Set Precision 67.37 # 4
Recall 69.12 # 5
F1-Measure 68.23 # 4
Named Entity Recognition (NER) Few-NERD (SUP) PIQN Precision 70.16 # 2
Recall 69.18 # 4
F1-Measure 69.67 # 2
Nested Named Entity Recognition GENIA PIQN F1 81.77 # 1
Chinese Named Entity Recognition MSRA PIQN F1 93.48 # 17
Nested Named Entity Recognition NNE PIQN Micro F1 94.04 # 3
Named Entity Recognition (NER) Ontonotes v5 (English) PIQN F1 90.96 # 6
Nested Named Entity Recognition TAC-KBP 2017 PIQN F1 84.5 # 3


No methods listed for this paper. Add relevant methods here