By the same authors

Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks

Research output: Working paper

Standard

Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks. / Wang, Yue; Xu, Zhuo; Bai, Lu; Wan, Yao; Cui, Lixin; Zhao, Qian; Hancock, Edwin R.; Yu, Philip S.

2020. (arXiv).

Research output: Working paper

Harvard

Wang, Y, Xu, Z, Bai, L, Wan, Y, Cui, L, Zhao, Q, Hancock, ER & Yu, PS 2020 'Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks' arXiv. <https://arxiv.org/abs/2010.06310>

APA

Wang, Y., Xu, Z., Bai, L., Wan, Y., Cui, L., Zhao, Q., Hancock, E. R., & Yu, P. S. (2020). Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks. (arXiv). https://arxiv.org/abs/2010.06310

Vancouver

Wang Y, Xu Z, Bai L, Wan Y, Cui L, Zhao Q et al. Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks. 2020 Oct 13. (arXiv).

Author

Wang, Yue ; Xu, Zhuo ; Bai, Lu ; Wan, Yao ; Cui, Lixin ; Zhao, Qian ; Hancock, Edwin R. ; Yu, Philip S. / Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks. 2020. (arXiv).

Bibtex - Download

@techreport{e0398d73f7d346f68b44e0318d451e61,
title = "Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks",
abstract = " Joint-event-extraction, which extracts structural information (i.e., entities or triggers of events) from unstructured real-world corpora, has attracted more and more research attention in natural language processing. Most existing works do not fully address the sparse co-occurrence relationships between entities and triggers, which loses this important information and thus deteriorates the extraction performance. To mitigate this issue, we first define the joint-event-extraction as a sequence-to-sequence labeling task with a tag set composed of tags of triggers and entities. Then, to incorporate the missing information in the aforementioned co-occurrence relationships, we propose a Cross-Supervised Mechanism (CSM) to alternately supervise the extraction of either triggers or entities based on the type distribution of each other. Moreover, since the connected entities and triggers naturally form a heterogeneous information network (HIN), we leverage the latent pattern along meta-paths for a given corpus to further improve the performance of our proposed method. To verify the effectiveness of our proposed method, we conduct extensive experiments on four real-world datasets as well as compare our method with state-of-the-art methods. Empirical results and analysis show that our approach outperforms the state-of-the-art methods in both entity and trigger extraction. ",
keywords = "cs.CL",
author = "Yue Wang and Zhuo Xu and Lu Bai and Yao Wan and Lixin Cui and Qian Zhao and Hancock, {Edwin R.} and Yu, {Philip S.}",
note = "Accepted by ICPR 2020",
year = "2020",
month = oct,
day = "13",
language = "Undefined/Unknown",
series = "arXiv",
type = "WorkingPaper",

}

RIS (suitable for import to EndNote) - Download

TY - UNPB

T1 - Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks

AU - Wang, Yue

AU - Xu, Zhuo

AU - Bai, Lu

AU - Wan, Yao

AU - Cui, Lixin

AU - Zhao, Qian

AU - Hancock, Edwin R.

AU - Yu, Philip S.

N1 - Accepted by ICPR 2020

PY - 2020/10/13

Y1 - 2020/10/13

N2 - Joint-event-extraction, which extracts structural information (i.e., entities or triggers of events) from unstructured real-world corpora, has attracted more and more research attention in natural language processing. Most existing works do not fully address the sparse co-occurrence relationships between entities and triggers, which loses this important information and thus deteriorates the extraction performance. To mitigate this issue, we first define the joint-event-extraction as a sequence-to-sequence labeling task with a tag set composed of tags of triggers and entities. Then, to incorporate the missing information in the aforementioned co-occurrence relationships, we propose a Cross-Supervised Mechanism (CSM) to alternately supervise the extraction of either triggers or entities based on the type distribution of each other. Moreover, since the connected entities and triggers naturally form a heterogeneous information network (HIN), we leverage the latent pattern along meta-paths for a given corpus to further improve the performance of our proposed method. To verify the effectiveness of our proposed method, we conduct extensive experiments on four real-world datasets as well as compare our method with state-of-the-art methods. Empirical results and analysis show that our approach outperforms the state-of-the-art methods in both entity and trigger extraction.

AB - Joint-event-extraction, which extracts structural information (i.e., entities or triggers of events) from unstructured real-world corpora, has attracted more and more research attention in natural language processing. Most existing works do not fully address the sparse co-occurrence relationships between entities and triggers, which loses this important information and thus deteriorates the extraction performance. To mitigate this issue, we first define the joint-event-extraction as a sequence-to-sequence labeling task with a tag set composed of tags of triggers and entities. Then, to incorporate the missing information in the aforementioned co-occurrence relationships, we propose a Cross-Supervised Mechanism (CSM) to alternately supervise the extraction of either triggers or entities based on the type distribution of each other. Moreover, since the connected entities and triggers naturally form a heterogeneous information network (HIN), we leverage the latent pattern along meta-paths for a given corpus to further improve the performance of our proposed method. To verify the effectiveness of our proposed method, we conduct extensive experiments on four real-world datasets as well as compare our method with state-of-the-art methods. Empirical results and analysis show that our approach outperforms the state-of-the-art methods in both entity and trigger extraction.

KW - cs.CL

M3 - Working paper

T3 - arXiv

BT - Cross-Supervised Joint-Event-Extraction with Heterogeneous Information Networks

ER -