Dual-Channel and Hierarchical Graph Convolutional Networks for document-level relation extraction

Sun, Qi; Xu, Tiancheng; Zhang, Kun; Huang, Kun; Lv, Laishui; Li, Xun; Zhang, Ting; Dore-Natteh, Doris

doi:10.1016/j.eswa.2022.117678

Cited by 17 publications

(4 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hypergraph neural networks (HGNNs) extend GNNs to hypergraphs and can learn node embeddings that capture both local and global information. [40][41][42] Hypergraphs are generalizations of graphs, where a hyperedge can connect any number of nodes, not just two. Another type of GNN is a hypergraph convolutional network (H-GCN) that performs spectral convolution on the hypergraph by using a hypergraph Laplacian.…”

Section: Convolutional Neural Network Model Based On Graph and Hyperg...mentioning

confidence: 99%

See 1 more Smart Citation

Pedestrian trajectory prediction with high-order interactions

Xiang,

Tian,

Wang

2024

J. Electron. Imag.

View full text Add to dashboard Cite

To tackle the problem of human trajectory prediction in complex scenes, we propose a model using hypergraph convolutional neural networks for social interaction (HGCNSI). Our model leverages a hypergraph structure to capture both high-order interactions and complex social dynamics among pedestrians (who often influence each other in a nonlinear and structured manner). First, we propose a social interaction module that improves the accuracy of interaction modeling by distinguishing between interacting and non-interacting pedestrians. Then, the hypergraph structure that can capture the complex and nonlinear relations among multiple pedestrians from the social interaction module is constructed. Furthermore, we exploit an improved attention mechanism called scene-coordinates attention that fuses the spatial and temporal features and models the unique historical movement information of each person. Finally, we introduce the SIRF module that filters the trajectories within one iteration to reduce the computational complexity and improve the prediction performance. We evaluate the proposed HGCNSI model on five publicly available datasets and demonstrate that it achieves state-of-the-art results. Specifically, the experiments show that our model outperforms existing methods in terms of prediction accuracy, using evaluation metrics, such as the average displacement error and the final displacement error.

show abstract

Section: Convolutional Neural Network Model Based On Graph and Hyperg...mentioning

confidence: 99%

“…Hypergraph neural networks (HGNNs) extend GNNs to hypergraphs and can learn node embeddings that capture both local and global information 40 – 42 Hypergraphs are generalizations of graphs, where a hyperedge can connect any number of nodes, not just two.…”

Section: Related Workmentioning

confidence: 99%

Pedestrian trajectory prediction with high-order interactions

Xiang,

Tian,

Wang

2024

J. Electron. Imag.

View full text Add to dashboard Cite

show abstract

“…However, in practical scenarios, entities within a text often exhibit their interrelations across multiple sentences, leading to instances of implicit relations. Although some studies have put forward methodologies for cross-sentence relation extraction (Sun et al, 2022a) to address these implicit relations, there is a need for further research to enhance the effectiveness of these existing approaches. The challenge becomes notably more complex when attempting to capture implicit relations that are not explicitly expressed within a single document but are dispersed throughout the extensive human knowledge that exists in written form.…”

Section: Explicit and Implicitmentioning

confidence: 99%

Towards the extraction of semantic relations in design with natural language processing

Giordano,

Consoloni,

Chiarello

et al. 2024

Proc. Des. Soc.

View full text Add to dashboard Cite

Natural Language Processing (NLP) has been extensively applied in design, particularly for analyzing technical documents like patents and scientific papers to identify entities such as functions, technical feature, and problems. However, there has been less focus on understanding semantic relations within literature, and a comprehensive definition of what constitutes a relation is still lacking. In this paper, we define relation in the context of design and the fundamental concepts linked to it. Subsequently, we introduce a framework for employing NLP to extract relations relevant to design.

show abstract

“…Zeng et al [24] constructed mention-level and entity-level document graphs, respectively, and they proposed a new path inference mechanism to infer relationships between entities. Sun et al [25] presented a dual-channel hierarchical graph convolutional neural network called DHGCN to model token-level, mention-level, and entity-level complex interactions between diferent semantics in a document. Transformerbased methods employ pretrained models (Bert, Roberta, ERNIE, etc.)…”

Section: Relation Extraction Relation Extraction Aims To Extract Rela...mentioning

confidence: 99%

A Novel Threat Intelligence Information Extraction System Combining Multiple Models

Guo

Chen

et al. 2022

Security and Communication Networks

View full text Add to dashboard Cite

The increasing number of cyberattacks has made the cybersecurity situation more serious. Thus, it is urgent to use cyber threat intelligence to deal with the complex and changing cyber environment. However, cyber threat intelligence usually exists in an unstructured form, and a huge amount of data poses a great challenge to security analysts. To this end, this paper proposes a novel threat intelligence information extraction system combining multiple models, which contains four key steps: entity extraction, coreference resolution, relation extraction, and knowledge graph construction. In the entity extraction task, a multihead self-attention mechanism is adopted to extract the dependency relationships between words. In the coreference resolution task, contextual information and mention embedding are fused to improve the mention representation. Meanwhile, features of different dimensions are extracted using a convolutional neural network. In the relation extraction task, additional features such as part of speech, mention width, entity type, and distance of entity pairs are incorporated to improve the embedding representation. Finally, a knowledge graph is constructed to explicitly present entities and their relationships. Experimental results indicate that compared with the baseline model, the F1 score of our model is improved by at least 8.87, 9.82, and 10.56 on entity extraction, coreference resolution, and relation extraction, respectively. The knowledge graph in Neo4j demonstrates the effectiveness of our system.

show abstract

Dual-Channel and Hierarchical Graph Convolutional Networks for document-level relation extraction

Cited by 17 publications

References 37 publications

Pedestrian trajectory prediction with high-order interactions

Pedestrian trajectory prediction with high-order interactions

Towards the extraction of semantic relations in design with natural language processing

A Novel Threat Intelligence Information Extraction System Combining Multiple Models

Contact Info

Product

Resources

About