Abstract
Event Extraction plays an important role in information-extraction to understand the world. Event extraction could be split into two subtasks: one is event trigger extraction, the other is event arguments extraction. However, the F-Score of event arguments extraction is much lower than that of event trigger extraction, i.e. in the most recent work, event trigger extraction achieves 80.7%, while event arguments extraction achieves only 58%. In pipelined structures, the difficulty of event arguments extraction lies in its lack of classification feature, and the much higher computation consumption. In this work, we proposed a novel Event Extraction approach based on multi-layer Dilate Gated Convolutional Neural Network (EE-DGCNN) which has fewer parameters. In addition, enhanced local information is incorporated into word features, to assign event arguments roles for triggers predicted by the first subtask. The numerical experiments demonstrated significant performance improvement beyond state-of-art event extraction approaches on real-world datasets. Further analysis of extraction procedure is presented, as well as experiments are conducted to analyze impact factors related to the performance improvement.
Highlights
The global web with electronic information, including most notably the WWW, provides a huge resource of unbounded information to understand the world
To improve the accuracy of assigning argument roles, we propose an elegant framework which has fewer parameters to model relationships among context in an order sensitive setting, as well as using the pre-trained representations with local features (The representations are obtained from large-scale open domain textual corpus, and the local features are obtained from the input)
In this paper, we propose a novel model for event arguments extraction based on multi-layer dilate gated convolutional neural network, which utilizing word embedding generated by the pre-trained model, as well as elaborately constructed token features
Summary
The global web with electronic information, including most notably the WWW, provides a huge resource of unbounded information to understand the world. Local features among event trigger and event arguments in an event mention are added to the proposed Dilate Gated Convolutional Neural Network based Event Extraction (EE-DGCNN) model input representation, which significantly improved the identification accuracy and classification accuracy of the extraction model. Both of them have achieved considerable performance and suffer from a huge number of model parameters These neural network-based methods usually combine the input word vector with a variety of sentence-level and word-level information and apply a carefully constructed neural network to capture word features. Since the Bidirectional Transformer model owns the ability to capture contextual information, which is significant for extracting mentions, we propose to complete those two sequence labeling tasks with representations generated by BERT. Word embeddings from a bidirectional transformer. Constructing token features representation with lexical features and contextual features. Assigning argument roles to each word in the sentence, with the predicted trigger
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
