TableIE: Capture the Interactions among Joint Information Extraction Explicitly via Double Tables
Information Extraction (IE) mainly consists of three sub-tasks, Named Entity Recognition, Relation Extraction, and Event Extraction. Although the sub-tasks are highly correlated with each other, most previous works simply focus on part of them and ignore the interactions among different sub-tasks. In this paper, we propose TableIE, a double-tables based method to capture the interactions among joint information extraction (Joint IE, i.e., conduct all sub-tasks jointly) explicitly. Different from the complicated graph-based Joint IE methods, we maintain two tables for entity-relation extraction and event extraction respectively instead and implement the interactions among three IE sub-tasks explicitly. The experiment show that our TableIE model outperforms the previous state-of-the-art up to 2.5 in the ACE05 dataset.
Code (0)
등록된 구현이 없습니다.
Tasks
Event Extractionnamed-entity-recognitionNamed Entity RecognitionNamed Entity Recognition (NER)RelationRelation ExtractionSimilar Papers 제목 키워드 기반
Information-Theoretic Measures on Lattices for High-Order Interactions
Traditional measures based solely on pairwise associations often fail to capture the complex statistical structure of multivariate data. Existing approaches for identifying information shared among $d>3$ variables are fr…
feature selectionModeling Dense Cross-Modal Interactions for Joint Entity-Relation Extraction
Joint extraction of entities and their relations benefits from the close interaction between named entities and their relation information. Therefore, how to effectively model such cross-modal interactions is critical fo…
Joint Entity and Relation ExtractionRelationRelation ClassificationRelation ExtractionCross-Modal Alignment via Variational Copula Modelling
Various data modalities are common in real-world applications (e.g., electronic health records, medical images and clinical notes in healthcare). It is essential to develop multimodal learning methods to aggregate variou…
IIP-Transformer: Intra-Inter-Part Transformer for Skeleton-Based Action Recognition
Recently, Transformer-based networks have shown great promise on skeleton-based action recognition tasks. The ability to capture global and local dependencies is the key to success while it also brings quadratic computat…
Action RecognitionData AugmentationSkeleton Based Action RecognitionJoint Event Extraction with Hierarchical Policy Network
Most existing work on event extraction (EE) either follows a pipelined manner or uses a joint structure but is pipelined in essence. As a result, these efforts fail to utilize information interactions among event trigger…
Event DetectionEvent Extraction