GapDNER: A Gap-Aware Grid Tagging Model for Discontinuous Named Entity Recognition

2025-10-12

AI Summary

Biomedical named entities often exhibit discontinuous and overlapping spans, posing challenges for existing methods that rely on span concatenation or token-level labeling—approaches prone to error propagation and decoding ambiguity. To address this, we propose a novel grid-tagging framework that, for the first time, explicitly models contextual gaps between entity fragments as independent semantic units. Our method introduces a bidirectional interaction module that jointly enhances fragment representations through two complementary pathways: intra-fragment regularity modeling (via biaffine and linear attention) and inter-fragment relational modeling (via cross-shaped attention). Furthermore, we adopt a BFS-based pathwise decoding strategy to ensure robust and consistent entity reconstruction. Evaluated on three benchmark biomedical NER datasets, our model achieves state-of-the-art performance, with particularly substantial gains in scenarios involving highly discontinuous and deeply nested entities.

Technology Category

• Natural Language Processing: Syntax — Tagging, Chunking & Parsing
• Reasoning under Uncertainty: Relational Probabilistic Models
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

In biomedical fields, one named entity may consist of a series of non-adjacent tokens and overlap with other entities. Previous methods recognize discontinuous entities by connecting entity fragments or internal tokens, which face challenges of error propagation and decoding ambiguity due to the wide variety of span or word combinations. To address these issues, we deeply explore discontinuous entity structures and propose an effective Gap-aware grid tagging model for Discontinuous Named Entity Recognition, named GapDNER. Our GapDNER innovatively applies representation learning on the context gaps between entity fragments to resolve decoding ambiguity and enhance discontinuous NER performance. Specifically, we treat the context gap as an additional type of span and convert span classification into a token-pair grid tagging task. Subsequently, we design two interactive components to comprehensively model token-pair grid features from both intra- and inter-span perspectives. The intra-span regularity extraction module employs the biaffine mechanism along with linear attention to capture the internal regularity of each span, while the inter-span relation enhancement module utilizes criss-cross attention to obtain semantic relations among different spans. At the inference stage of entity decoding, we assign a directed edge to each entity fragment and context gap, then use the BFS algorithm to search for all valid paths from the head to tail of grids with entity tags. Experimental results on three datasets demonstrate that our GapDNER achieves new state-of-the-art performance on discontinuous NER and exhibits remarkable advantages in recognizing complex entity structures.

Problem

• Recognizing discontinuous named entities with non-adjacent tokens
• Resolving decoding ambiguity in discontinuous entity recognition
• Modeling context gaps between entity fragments for improved performance

Innovation

• Models context gaps between entity fragments for NER
• Uses token-pair grid tagging with intra-span and inter-span modules
• Employs BFS algorithm for decoding valid entity paths