Document-level Relation Extraction with Dual-tier Heterogeneous Graph

Zhang, Zhenyu; Yu, Bowen; Shu, Xiaobo; Liu, Tingwen; Tang, Hengzhu; Wang, Yubin; Guo, Li

doi:10.18653/v1/2020.coling-main.143

Cited by 42 publications

(16 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In experiments on two biomedical datasets, we compared SAM-LSTM with many models, including EoG [6], LSR [7], DHG [17], GLRE [9], and ATLOP [14]. We applied the SciBERTbase [31] model pre-trained in the scientific publications corpus.…”

Section: Results On the Biomedical Datasetsmentioning

confidence: 99%

“…The authors of [16] proposed a graphically enhanced dual attention network (GEDA) for attentional supervision from additional evidence to construct complex interactions between sentences and potential relational instances. The authors of [17] proposed a Dual-tier Heterogeneous Graph (DHG) to model the structural information of documents and perform relational reasoning across sentences. The authors of [18] proposed an encoder-classifier reconstructor model (HeterGSAN) that reconstructs the paths and puts more attention on related entity pairs instead of negative samples.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A Keyword Detection and Context Filtering Method for Document Level Relation Extraction

Kuang

Chen

et al. 2022

Applied Sciences

View full text Add to dashboard Cite

Relation extraction (RE) is the core link of downstream tasks, such as information retrieval, question answering systems, and knowledge graphs. Most of the current mainstream RE technologies focus on the sentence-level corpus, which has great limitations in practical applications. Moreover, the previously proposed models based on graph neural networks or transformers try to obtain context features from the global text, ignoring the importance of local features. In practice, the relation between entity pairs can usually be inferred just through a few keywords. This paper proposes a keyword detection and context filtering method based on the Self-Attention mechanism for document-level RE. In addition, a Self-Attention Memory (SAM) module in ConvLSTM is introduced to process the document context and capture keyword features. By searching for word embeddings with high cross-attention of entity pairs, we update and record critical local features to enhance the performance of the final classification model. The experimental results on three benchmark datasets (DocRED, CDR, and GBA) show that our model achieves advanced performance within open and specialized domain relationship extraction tasks, with up to 0.87% F1 value improvement compared to the state-of-the-art methods. We have also designed experiments to demonstrate that our model can achieve superior results by its stronger contextual filtering capability compared to other methods.

show abstract

Section: Results On the Biomedical Datasetsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

A Keyword Detection and Context Filtering Method for Document Level Relation Extraction

Kuang

Chen

et al. 2022

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…On the CDR and GDA data sets, we compared BRAN (Verga et al, 2018), EoG (Christopoulou et al, 2019), LSR (Nan et al, 2020), DHG (Zhang et al, 2020c), GLRE (Wang et al, 2020a), ATLOP (Zhou et al, 2021), and DocuNet with our model. The experimental results on two biomedical datasets are shown in Table 2.…”

Section: Results On the Biomedical Datasetsmentioning

confidence: 99%

“…LSR (Nan et al, 2020) 64.8 82.2 DHG (Zhang et al, 2020c) 65.9 83.1 GLRE (Wang et al, 2020a) 68.5 -SciBERT base (Beltagy et al, 2019) 65.1 82.5 ATLOP-SciBERT base (Zhou et al, 2021) 69.4 83.9…”

Section: Ablation Studymentioning

confidence: 99%

A Densely Connected Criss-Cross Attention Network for Document-level Relation Extraction

Zhang¹,

Cheng²

2022

Preprint

View full text Add to dashboard Cite

Document-level relation extraction (RE) aimsto identify relations between two entities in a given document.Compared with its sentence-level counterpart, document-level RE requires complex reasoning. Previous research normally completed reasoning through information propagation on the mention-level or entity-level document-graph, but rarely considered reasoning at the entity-pair-level. In this paper, we propose a novel model, called Densely Connected Criss-Cross Attention Network (Dense-CCNet), for documentlevel RE, which can complete logical reasoning at the entity-pair-level. Specifically, the Dense-CCNet performs entity-pair-level logical reasoning through the Criss-Cross Attention (CCA), which can collect contextual information in horizontal and vertical directions on the entity-pair matrix to enhance the corresponding entity-pair representation. In addition, we densely connect multiple layers of the CCA to simultaneously capture the features of single-hop and multi-hop logical reasoning. We evaluate our Dense-CCNet model on three public document-level RE datasets, DocRED, CDR, and GDA. Experimental results demonstrate that our model achieves state-of-the-art performance on these three datasets.

show abstract

“…Chaojun Xiao et al [20]used Bert to encode document structure information for Document-level intersentence relation extraction; Benfeng Xu et al [21] proposed SSAN model based on Transformer, which incorporated graph information into the In recent years, graph-based approaches have gradually become the mainstream technique for document-level relation extraction due to the advantages of graph neural networks in mining document structure features. Zhenyu Zhang et al [24] proposed a two-layer heterogeneous graph that separates document structure modeling layer from relation inference layer; Guoshun Nan et al [16] proposed a LSP model, which used syntactic dependencies to construct graph. Graph nodes automatically proceed to learn more non-neighbor information through neighbor nodes in a fully connected state.…”

Section: Related Workmentioning

confidence: 99%

Inter-sentence Entity Relation Extraction based on GNN of Message Propagation

Chen

Chunxiang

Guo

et al. 2022

Preprint

View full text Add to dashboard Cite

Compared with sentence-level relation extraction, it is more common that entities involving relationships exist in multiple sentences in actual scenarios. Therefore, document-level relation extraction has gradually become a research hotspot in the field of information extraction in recent years. In order to make full use of the contextual information and structure information of the document, this paper combine embedding representation of entities and the explicit structural relationships between entities to study extraction of document-level inter-sentence relations. Firstly, document is encoded using BiLSTM, while document graph is established using the explicit structure of document; Secondly, relational graph convolutional neural network based on message propagation is used to update nodes in the graph, thus local information of document is integrated into node embedding representation, using which edge embedding is updated; Finally, an iterative algorithm is used to complete the reasoning of edge information, and entity relation is predicted using a feed-forward network based on edge embedding. Experimental results show that, compared with the EoG model, the F1 values of inter-sentence relation are increased by 2.6% and 0.8% on the CDR and GDA datasets, respectively.

show abstract

Document-level Relation Extraction with Dual-tier Heterogeneous Graph

Cited by 42 publications

References 32 publications

A Keyword Detection and Context Filtering Method for Document Level Relation Extraction

A Keyword Detection and Context Filtering Method for Document Level Relation Extraction

A Densely Connected Criss-Cross Attention Network for Document-level Relation Extraction

Inter-sentence Entity Relation Extraction based on GNN of Message Propagation

Contact Info

Product

Resources

About