Applicability Analysis and Ensemble Application of BERT with TF-IDF, TextRank, MMR, and LDA for Topic Classification Based on Flood-Related VGI DOI Creative Commons
Wenying Du, Chang Ge, Shuang Yao

et al.

ISPRS International Journal of Geo-Information, Journal Year: 2023, Volume and Issue: 12(6), P. 240 - 240

Published: June 9, 2023

Volunteered geographic information (VGI) plays an increasingly crucial role in flash floods. However, topic classification and spatiotemporal analysis are complicated by the various expressions lengths of social media textual data. This paper conducted applicability on bidirectional encoder representation from transformers (BERT) four traditional methods, TextRank, term frequency–inverse document frequency (TF-IDF), maximal marginal relevance (MMR), linear discriminant (LDA), results show that for user type, BERT performs best Government Affairs Microblog, whereas LDA-BERT We Media Microblog. As text length, TF-IDF-BERT works better texts with a length <70 >140 words, 70–140 words. For evolution pattern, study suggests Henan rainstorm, topics follow general pattern “situation-tips-rescue”. Moreover, this detected hotspot “Metro Line 5” related to rainstorm discovered topical focus spatially shifts Zhengzhou, first Xinxiang, then Hebi, showing remarkable tendency south north, which was same as report issued authorities. integrated multi-methods improve overall accuracy Sina microblogs, facilitating flooding.

Language: Английский

The influence of road network topology on street flooding in New York City—A social media data approach DOI
Chen Zuo, Runzi Wang, Yi Hong

et al.

Journal of Hydrology, Journal Year: 2024, Volume and Issue: 638, P. 131471 - 131471

Published: June 16, 2024

Language: Английский

Citations

5
Effects of urban green spaces landscape pattern on carbon sink among urban ecological function areas at the appropriate scale: A case study in Xi’an DOI Creative Commons

Nuo Shi,

Yu Yang,

Shuting Liang

et al.

Ecological Indicators, Journal Year: 2023, Volume and Issue: 158, P. 111427 - 111427

Published: Dec. 20, 2023

At appropriate scales, optimizing landscape pattern in urban green spaces is shown to augment carbon sink. However, existing research primarily concentrates on regional or city-wide areas, often overlooking the nuanced effects of sink within specialized ecological function areas. Focusing city Xi'an and emphasizing forest as principal subject, this study investigated spatial interplay between across Ecological Control Area, Urban Construction Buffer Tsinling Mountains Protection Area. Grounded field surveys, employed GF1-WFV remote sensing imagery ascertain 2021 land-use data further classified forests into five distinct types: deciduous broad-leaved forest, evergreen coniferous mixed evergreen-deciduous forest. Employing Carnegie-Ames-Stanford Approach model metrics methods, evaluated at multi-scale. Through semi-variance analyses, we discerned grid scales. Finally, multiple linear stepwise regression elucidated significant metrics, while geographically weighted were scrutinize relationships. Our findings suggested that scales 5 km for 3 6 Area are most effective management planning. Metrics related area, edge, fragmentation pivotal enhancing sink, their influence was spatially variable. This furnished an in-depth analysis how interact various areas thereby providing a scientific framework sustainable planning spaces.

Language: Английский

Citations

13
An integrated framework for flood disaster information extraction and analysis leveraging social media data: A case study of the Shouguang flood in China DOI

Hong-bing Hou,

Li Shen,

Jianan Jia

et al.

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 949, P. 174948 - 174948

Published: July 25, 2024

Language: Английский

Citations

4
Frontiers of policy and governance research in a smart city and artificial intelligence: an advanced review based on natural language processing DOI Creative Commons
Liang Dong, Yunhong Liu

Frontiers in Sustainable Cities, Journal Year: 2023, Volume and Issue: 5

Published: July 27, 2023

This study presents an advanced review of policy and governance research in the context smart cities artificial intelligence (AI). With playing a crucial role achieving United Nations Sustainable Development Goals, it is vital to understand opportunities challenges that arise from applications technologies AI promoting urban sustainability. Using Latent Dirichlet Allocation (LDA) method based on three-layer Bayesian algorithm model, we conducted systematic approximately 3700 papers Scopus. Our analysis revealed prominent topics such as “service transformation,” “community participation,” “sustainable development goals.” We also identified emerging concerns, including “open user data,” “ethics risk management,” “data privacy management.” These findings provide valuable insights into current progress frontiers field, informing future directions decision-making processes.

Language: Английский

Citations

11
Applicability Analysis and Ensemble Application of BERT with TF-IDF, TextRank, MMR, and LDA for Topic Classification Based on Flood-Related VGI DOI Creative Commons
Wenying Du, Chang Ge, Shuang Yao

et al.

ISPRS International Journal of Geo-Information, Journal Year: 2023, Volume and Issue: 12(6), P. 240 - 240

Published: June 9, 2023

Volunteered geographic information (VGI) plays an increasingly crucial role in flash floods. However, topic classification and spatiotemporal analysis are complicated by the various expressions lengths of social media textual data. This paper conducted applicability on bidirectional encoder representation from transformers (BERT) four traditional methods, TextRank, term frequency–inverse document frequency (TF-IDF), maximal marginal relevance (MMR), linear discriminant (LDA), results show that for user type, BERT performs best Government Affairs Microblog, whereas LDA-BERT We Media Microblog. As text length, TF-IDF-BERT works better texts with a length <70 >140 words, 70–140 words. For evolution pattern, study suggests Henan rainstorm, topics follow general pattern “situation-tips-rescue”. Moreover, this detected hotspot “Metro Line 5” related to rainstorm discovered topical focus spatially shifts Zhengzhou, first Xinxiang, then Hebi, showing remarkable tendency south north, which was same as report issued authorities. integrated multi-methods improve overall accuracy Sina microblogs, facilitating flooding.

Language: Английский

Citations

10