Xiaoqi Shen

Dr. Xiaoqi Shen

orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 May 2022 in Humanities and Social Sciences Communications
Verified
Journal Article
Open access
PDF file
WSWen-Zhong Shi
Wen-Zhong Shi
FZFanxin Zeng
Fanxin Zeng
AZAnshu Zhang
Anshu Zhang
CTChengzhuo Tong
Chengzhuo Tong
XSXiaoqi Shen
Xiaoqi Shen
ZSZhicheng Shi
Zhicheng Shi
As COVID-19 spread around the world, epidemic prevention and control policies have been adopted by many countries. This process has prompted online social platforms to become important channels to enable people to socialize and exchange information. The massive use of social media data mining techniques, to analyze the development online of public opinion during the epidemic, is of great significance in relation to the management of public opinion. This paper presents a study that aims to analyze the developmental course of online public opinion in terms of fine-grained emotions presented during the COVID-19 epidemic in China. It is based on more than 45 million Weibo posts during the period from December 1, 2019 to April 30, 2020. A text emotion extraction method based on a dictionary of emotional ontology has been developed. The results show, for example, that a high emotional effect is observed during holidays, such as New Year. As revealed by Internet users, the outbreak of the COVID-19 epidemic and its rapid spread, over a comparatively short period of time, triggered a sharp rise in the emotion “fear”. This phenomenon was noted especially in Wuhan and the immediate surrounding areas. Over the initial 2 months, although this “fear” gradually declined, it remained significantly higher than the more common level of uncertainty that existed during the epidemic’s initial developmental era. Simultaneously, in the main city clusters, the response to the COVID-19 epidemic in central cities, was stronger than that in neighboring cities, in terms of the above emotion. The topics of Weibo posts, the corresponding emotions, and the analysis conclusions can provide auxiliary reference materials for the monitoring of network public opinion under similar major public events.

13 November 2020 in International Journal of Geographical Information Science
Verified
Journal Article
Closed access
Zhewei LIUZhewei LIU
Zhewei LIU
AZAnshu Zhang
Anshu Zhang
YYYepeng Yao
Yepeng Yao
Wenzhong ShiWenzhong Shi
Wenzhong Shi
Xiao HuangXiao Huang
Xiao Huang
XSXiaoqi Shen
Xiaoqi Shen
Various methods have been proposed to detect the base locations of individuals, with their geo-tagged social media data. However, a common challenge relating to base-location detection methods (BDMs) is that, the rare availability of ground-truth data impedes the method assessment of accuracy and robustness, thus undermining research validity and reliability. To address this challenge, we collect users’ information from unstructured online content, and evaluate both the performance and robustness of BDMs. The evaluation consists of two tasks: the detection of base locations and also the differentiation between local residents and tourists. The results show BDMs can achieve high accuracies in base-location detection but tend to overestimate the number of tourists. Evaluation conducted in this study, also shows that BDMs’ accuracy is subject to the intensity of user’s activities and number of countries visited by the user but are insensitive to user’s gender. Temporally, BDMs perform better during weekends and summertime than during other periods, but the best performances appear with datasets that cover the whole time periods (whole day, week, and year). To the best of knowledge, this study is the first work to evaluate the performance and robustness of BDMs at individual level.

15 February 2023 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
LWLukang Wang
Lukang Wang
Min ZhangMin Zhang
Min Zhang
XSXiaoqi Shen
Xiaoqi Shen
Wenzhong ShiWenzhong Shi
Wenzhong Shi
Landslide mapping (LM) from bitemporal remote sensing images is essential for disaster prevention and mitigation. Bitemporal change detection technology has been applied for LM, however, there remains room for improvement in its accuracy and automation. In this paper, a multi-level feature enhancement network (MFENet) is proposed for LM based on modules built in convolutional neural network (CNN) like CNN-Attention. MFENet mainly consists of three modules: post-event feature enhancement module (PFEM), bi-feature difference enhancement module (BFDEM) and flow direction calibration module (FDCM). Specifically, the main role of PFEM selectively fuses post-event multi-layer features to provide discriminative post-event features. BFDEM fuses the multi-layer differences of both pre-event and post-event features to generate high-quality change detection features, sufficiently powerful to distinguish foreground from background. FDCM uses a digital elevation model (DEM) to calibrate the flow direction of each pixel of the landslide detection results to complete the LM task. Experiments were conducted to test the effectiveness of MFENet on two real-world regions, Lantau Island and Sharp Peak, Hong Kong, in which landslides occurred after rainstorms. Compared with other state-of-the-art general change detection methods and landslide-specific change detection methods, the proposed method outperforms on all metrics with its intersection over union (IoU) reaching 87.23%. The availability of additional features and the generalization performance of MFENet are demonstrated experimentally. It is anticipated that the proposed network will further contribute to disaster prevention and mitigation.

15 August 2023 in IEEE Geoscience and Remote Sensing Letters
Verified
Journal Article
Closed access
LWLukang Wang
Lukang Wang
YLYue Li
Yue Li
Min ZhangMin Zhang
Min Zhang
XSXiaoqi Shen
Xiaoqi Shen
WPWenguang Peng
Wenguang Peng
Wenzhong ShiWenzhong Shi
Wenzhong Shi
Change detection (CD) is an important application of remote sensing (RS) technology, which discovers changes by observing bi-temporal RS images. The rise of deep learning provides new solutions for CD. However, due to the insufficient extraction and utilization of deep features from RS images, existing deep learning-based CD methods are difficult to fully integrate such deep features, resulting in unstable performance, especially low sensitivity to multi-scale changes. In this letter, a multi-scale feature fusion CD network (MSFF-CDNet) is proposed to enhance feature fusion, by integrating a mask guided change fusion module (MGCF) to achieve the fusion of the consistency and difference of multi-scale features. Also, a CD refinement module (CDRM) is implemented to assist the encoding-decoding structure to achieve CD at a finer scale. By training with a hybrid loss function, the MSFF-CDNet is able to learn trans-formation relationships of bi-temporal RS images and their change maps. Besides, using a deep supervised learning strategy further improves the fitting performance and robustness. The method is experimented on two open-source datasets (i.e., CDD and LEVIR-CD dataset). Compared to state-of-the-art CD methods, the proposed method outperforms on all metrics and its IoU reaches 92.39% and 85.89%, respectively. The codes are available at https://github.com/WangLukang/MSCD.

13 January 2022 in GIScience & Remote Sensing
Verified
Journal Article
Open access
XSXiaoqi Shen
Xiaoqi Shen
Wenzhong ShiWenzhong Shi
Wenzhong Shi
PCPengfei CHEN
Pengfei CHEN
Zhewei LIUZhewei LIU
Zhewei LIU
LWLukang Wang
Lukang Wang
The increasing amount of geotagged social media data provides a possible resource for location prediction. However, existing location prediction methods rarely incorporate temporal changes in mobility patterns, which could lead to unreliable predictions. In particular, human mobility patterns have changed greatly in the COVID-19 era. We propose a novel model to predict individuals’ movements in dynamic regions of interest (ROIs), taking into account changes in activity areas and movement regularity. To address changes in the activity areas, we design a new updating strategy that can ensure the realistic extraction of an individual’s ROIs. Then, we develop an integration model for changes in the movement regularity based on two newly proposed prediction methods that consider both rapid and slow changes. The proposed integration model is evaluated based on five real-world social media datasets; three Weibo datasets related to COVID-19 collected in three Chinese cities, one Twitter dataset collected in New York and one dense GPS dataset. The results demonstrate that the proposed model can achieve better performances than state-of-the-art models, especially when mobility patterns change greatly. Combined with related pandemic data, this study will benefit pandemic prevention and control.

6 October 2020 in International Journal of Geographical Information Science
Verified
Journal Article
Closed access
Wenzhong ShiWenzhong Shi
Wenzhong Shi
PCPengfei CHEN
Pengfei CHEN
XSXiaoqi Shen
Xiaoqi Shen
JLJianxiao Liu
Jianxiao Liu
Modelling movement uncertainty is of profound significance in promoting effective trajectory analysis and mining. However, classic uncertainty models are limited by rigid assumptions on moving speed and distance, which ignores the stochastic nature of individual’s travel behaviour. This study introduces a novel method using adaptive ellipses to represent the movement uncertainty in a planar space under the framework of time geography. Two models are established by considering different error sources in trajectory data. The first model captures the uncertainty caused by sampling error, and the second one further, takes the measurement error into account. The Minkowski distance metric is adopted to determine the size of uncertainty ellipses, in which the Minkowski parameter is optimized for each segment in the raw trajectory on the basis of the geometric characteristics extracted. Compared with state-of-the-art methods on five real-life trajectory datasets, the proposed method is proved to produce more effective uncertain regions, which significantly reduce the redundant uncertain area, while retaining at a comparative level of actual movement coverage. Given the heterogeneity of human mobility patterns, this study provides a robust and applicable solution for adaptively modelling individual’s movement uncertainty, which is expected to benefit trajectory-related applications in various scenarios.