Search

Your search keyword '"Gu Xiaolei"' showing total 23 results
Start Over Author "Gu Xiaolei" Publication Year Range Last 3 years
23 results on '"Gu Xiaolei"'

1. Bipartite Graph Pre-training for Unsupervised Extractive Summarization with Graph Convolutional Auto-Encoders

Author

Mao, Qianren, Zhao, Shaobo, Li, Jiarui, Gu, Xiaolei, He, Shizhu, Li, Bo, and Li, Jianxin

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence
Abstract

Pre-trained sentence representations are crucial for identifying significant sentences in unsupervised document extractive summarization. However, the traditional two-step paradigm of pre-training and sentence-ranking, creates a gap due to differing optimization objectives. To address this issue, we argue that utilizing pre-trained embeddings derived from a process specifically designed to optimize cohensive and distinctive sentence representations helps rank significant sentences. To do so, we propose a novel graph pre-training auto-encoder to obtain sentence embeddings by explicitly modelling intra-sentential distinctive features and inter-sentential cohesive features through sentence-word bipartite graphs. These pre-trained sentence representations are then utilized in a graph-based ranking algorithm for unsupervised summarization. Our method produces predominant performance for unsupervised summarization frameworks by providing summary-worthy sentence representations. It surpasses heavy BERT- or RoBERTa-based sentence representations in downstream tasks., Comment: Accepted by the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023)

Published
2023

4. A Study on the Selection Mechanism of the Chinese Loan Words of the Same Signified with Different Signifiers

Author

Chen, Wenqi, Gan, Jiaming, Gu, Xiaolei, Li, Jin, Wang, Ruozhu, Goos, Gerhard, Series Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Su, Qi, editor, Xu, Ge, editor, and Yang, Xiaoyan, editor

Published
2023
Full Text
View/download PDF

16. Post-buckling behavior and failure analysis of asymmetric sandwich panels under uniaxial compression.

Author

Wu, Dake, Gu, Xiaolei, Peng, Ang, Deng, Jian, Cai, Deng'an, Zhou, Guangming, and Wang, Xinwei

Subjects
*SANDWICH construction (Materials), *FAILURE analysis, *BEHAVIORAL assessment, *STRESS concentration
Abstract

Sandwich structures with an asymmetric configuration are employed to introduce in-plane loads as well as to connect them with other structural components. Understanding their failure mechanism is crucial to ensure the safety design of aircraft. This paper investigates experimentally and numerically the mechanical behavior of asymmetric sandwich panels (ASPs) under compression. Quasi-static tests are conducted to obtain the buckling loads, post-buckling behavior, and load-bearing capacity of ASPs. A non-linear finite element (FE) model is developed to elucidate their damage behavior and failure mechanism. Numerical predictions are compared with test results to validate the FE model and failure analysis. The results demonstrate that the presence of a tapered region significantly influences stress distribution and failure initiations. A parametric study is also conducted to investigate the effects of the tapered angle on the compression response of ASPs. The reported results can provide a reference for the optimal design of ASPs. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

17. HSNF: Hybrid sampling with two-step noise filtering for imbalanced data classification.

Author

Duan, Lilong, Xue, Wei, Gu, Xiaolei, Luo, Xiao, and He, Yongsheng

Subjects
EUCLIDEAN distance, MACHINE learning, CLASSIFICATION, PROBLEM solving, NOISE
Abstract

Imbalanced data classification has received much attention in machine learning, and many oversampling methods exist to solve this problem. However, these methods may suffer from insufficient noise filtering, overlap between synthetic and original samples, etc., resulting in degradation of classification performance. To this end, we propose a hybrid sampling with two-step noise filtering (HSNF) method in this paper, which consists of three modules. In the first module, HSNF denoises twice according to different noise discrimination mechanisms. Note that denoising mechanism is essentially based on the Euclidean distance between samples. Then in the second module, the minority class samples are divided into two categories, boundary samples and safe samples, respectively, and a portion of the boundary majority class samples are removed. In the third module, different oversampling methods are used to synthesize instances for boundary minority class samples and safe minority class samples. Experimental results on synthetic data and benchmark datasets demonstrate the effectiveness of HSNF in comparison with several popular methods. The code of HSNF will be released. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

20. The status of surrogacy in China

Author

Qi, Qing, primary, Gu, Xiaolei, additional, Zhao, Yangyang, additional, Chen, Ziqin, additional, Zhou, Jing, additional, Chen, Song, additional, and Wang, Ling, additional

Published
2023
Full Text
View/download PDF

23. Potential Anti-Depressive Effects and Mechanisms of Zhi-Zi Hou-Po Decoction Using Behavioral Despair Tests Combined With in Vitro Approaches

Author

Bai, Yongtao, Dai, Guoliang, Song, Lihua, Gu, Xiaolei, Ba, Ning, Ju, Wenzheng, and Zhang, Wenzhou

Subjects
Pharmacology, Pharmacology (medical)
Abstract

Zhi-Zi Hou-Po Decoction (ZHD) has been widely used in the treatment of depression for centuries. This study aimed to investigate the antidepressant effects of the water extract of ZHD (ZHD-WE) and ethanol extract of ZHD (ZHD-EE) using behavioral despair tests in mice, and to further explore the neuroprotective effects in a PC12 cell injury model induced by corticosterone (CORT). Mice were divided into a control group (normal saline), ZHD-WE groups (4, 8, and 16 g kg−1), ZHD-EE groups (4, 8, and 16 g kg−1) and the fluoxetine group (20 mg kg−1). The forced swimming test (FST) and tail suspension test (TST) were used to screen the antidepressant effects of ZHD-WE and ZHD-EE after oral administration for seven consecutive days. The level of brain-derived neurotrophic factor (BDNF) in the hippocampus was determined by ELISA. The MTT, lactate dehydrogenase (LDH) and flow cytometry analysis were performed to elucidate the neuroprotective effect of ZHD-EE on a PC12 cell injury model. Additionally, the mRNA and proteins expression of apoptotic molecules Bax, Bcl-2 and BDNF were detected by RT-PCR and western blot assay. It showed that ZHD-EE at concentrations of 8 and 16 g kg−1 significantly decreased the immobility time in the TST and FST, and increased the BDNF levels in the hippocampus. While ZHD-WE at concentrations of 4, 8, and 16 g kg−1 had no significant effect on the immobility time in the TST, and only the 16 g kg−1 of extract group significantly decreased the immobility time in the FST. In vitro, the obtained results showed that PC12 cells pre-incubated with ZHD-EE at concentrations of 100 and 400 μg ml−1 improved cell viability, decreased LDH release, and reduced apoptosis rate of PC12 cells. Moreover, ZHD-EE significantly increased the mRNA and proteins expression of Bcl-2 and BDNF, while decreased the mRNA and protein expression of Bax. ZHD-EE significantly improved despair-like behavior in mice, and its mechanism may be related to BDNF upregulation in the hippocampus. This study also showed that ZHD-EE had a protective effect on CORT-induced injury in PC12 cells by upregulating the expression of BDNF and restoring Bcl-2/Bax balance.

Published
2022

Catalog

Books, media, physical & digital resources
See catalog results