Social Media Text Mining and Flood Disaster Analysis of Small Towns in Southern Shaanxi Qinba Mountain Area Based on Deep Learning
| dc.contributor.author | Zhao, Xin | |
| dc.contributor.author | Wu, Zuobin | |
| dc.date.accessioned | 2024-11-26T07:03:29Z | |
| dc.date.available | 2024-11-26T07:03:29Z | |
| dc.date.issued | 2024 | en |
| dc.description | Game changer? Planning for just and sustainable urban regions, Paris, 8-12th July 2024 | en |
| dc.description.abstract | The advent of the dataization era has made social media a new trend and tool for analyzing and managing flood risk. This paper aims to use the BERT-BiLSTM-CRF method to analyze the information on social media such as Weibo, and extract the content related to the flood of small towns in the southern Shaanxi Qinba mountain area. Firstly, we use python crawler to crawl the text data on social media such as Weibo, and then preprocess the data, including removing stop words, punctuation marks, emoticons, etc. Secondly, we use the BERT-BiLSTM-CRF method to perform named entity recognition on the text data, identify entities such as place names, person names, organization names, etc., and annotate them in BIO format. Then, we use methods such as geodetector to geocode the identified place name entities, obtain their latitude and longitude coordinates, and match them with the flood data of small towns in the southern Shaanxi Qinba mountain area, and analyze their flood distribution and flood relationship. Finally, we use flood risk analysis methods, such as flood frequency analysis, flood depth analysis, flood loss analysis, etc., to evaluate the similarity and difference of Weibo information and urban flood risk, and explore the impact and reflection of Weibo information on urban flood risk. In addition, using the high-precision risk analysis method of social media analysis, the flood risk and danger of small towns in the southern Shaanxi Qinba mountain area are evaluated, and the potential impact of flood disaster on the spatial characteristics of small towns is analyzed. The innovation of this paper lies in the combination of deep learning methods and flood analysis methods, extracting the information related to small town floods from large-scale social media data, providing new data sources and analysis methods for towns flood prevention and disaster reduction, and considering the impact factors of flood disaster, providing reference for small town planning and development. Keywords: Flood Disaster Analysis, climate change, Flood resilience strategies, Small Town, Planning, Social Media Text Mining | |
| dc.description.version | published version | en |
| dc.identifier.isbn | 978-94-64981-82-7 | en |
| dc.identifier.pageNumber | 3384-3394 | |
| dc.identifier.uri | https://hdl.handle.net/20.500.14235/2164 | |
| dc.language.iso | English | en |
| dc.publisher | AESOP | en |
| dc.rights | openAccess | en |
| dc.rights.license | CC-BY | en |
| dc.source | Game changer? Planning for just and sustainable urban regions, Paris, 8-12th July 2024 | en |
| dc.title | Social Media Text Mining and Flood Disaster Analysis of Small Towns in Southern Shaanxi Qinba Mountain Area Based on Deep Learning | |
| dc.type | conferenceObject | en |
| dc.type.version | publishedVersion | en |