Dynamic Spatio-Temporal Tweet Mining for Event Detection : A Case Study of Hurricane Florence
(2020) In International Journal of Disaster Risk Science 11(3). p.378-393- Abstract
Extracting information about emerging events in large study areas through spatiotemporal and textual analysis of geotagged tweets provides the possibility of monitoring the current state of a disaster. This study proposes dynamic spatio-temporal tweet mining as a method for dynamic event extraction from geotagged tweets in large study areas. It introduces the use of a modified version of ordering points to identify the clustering structure to address the intrinsic heterogeneity of Twitter data. To precisely calculate the textual similarity, three state-of-the-art text embedding methods of Word2vec, GloVe, and FastText were used to capture both syntactic and semantic similarities. The impact of selected embedding algorithms on the... (More)
Extracting information about emerging events in large study areas through spatiotemporal and textual analysis of geotagged tweets provides the possibility of monitoring the current state of a disaster. This study proposes dynamic spatio-temporal tweet mining as a method for dynamic event extraction from geotagged tweets in large study areas. It introduces the use of a modified version of ordering points to identify the clustering structure to address the intrinsic heterogeneity of Twitter data. To precisely calculate the textual similarity, three state-of-the-art text embedding methods of Word2vec, GloVe, and FastText were used to capture both syntactic and semantic similarities. The impact of selected embedding algorithms on the quality of the outputs was studied. Different combinations of spatial and temporal distances with the textual similarity measure were investigated to improve the event detection outcomes. The proposed method was applied to a case study related to 2018 Hurricane Florence. The method was able to precisely identify events of varied sizes and densities before, during, and after the hurricane. The feasibility of the proposed method was qualitatively evaluated using the Silhouette coefficient and qualitatively discussed. The proposed method was also compared to an implementation based on the standard density-based spatial clustering of applications with noise algorithm, where it showed more promising results.
(Less)
- author
- Farnaghi, Mahdi LU ; Ghaemi, Zeinab and Mansourian, Ali LU
- organization
- publishing date
- 2020-06
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Disaster management, Hurricane Florence, Natural language processing, Spatio-temporal tweet analysis, Tweet clustering, Twitter, Machine Learning (ML)
- in
- International Journal of Disaster Risk Science
- volume
- 11
- issue
- 3
- pages
- 16 pages
- publisher
- Springer
- external identifiers
-
- scopus:85085495747
- ISSN
- 2095-0055
- DOI
- 10.1007/s13753-020-00280-z
- language
- English
- LU publication?
- yes
- id
- 22f834cd-4579-4e97-a7d8-9491eadc3e75
- date added to LUP
- 2020-06-08 11:34:16
- date last changed
- 2023-10-08 05:37:27
@article{22f834cd-4579-4e97-a7d8-9491eadc3e75,
abstract = {{<p>Extracting information about emerging events in large study areas through spatiotemporal and textual analysis of geotagged tweets provides the possibility of monitoring the current state of a disaster. This study proposes dynamic spatio-temporal tweet mining as a method for dynamic event extraction from geotagged tweets in large study areas. It introduces the use of a modified version of ordering points to identify the clustering structure to address the intrinsic heterogeneity of Twitter data. To precisely calculate the textual similarity, three state-of-the-art text embedding methods of Word2vec, GloVe, and FastText were used to capture both syntactic and semantic similarities. The impact of selected embedding algorithms on the quality of the outputs was studied. Different combinations of spatial and temporal distances with the textual similarity measure were investigated to improve the event detection outcomes. The proposed method was applied to a case study related to 2018 Hurricane Florence. The method was able to precisely identify events of varied sizes and densities before, during, and after the hurricane. The feasibility of the proposed method was qualitatively evaluated using the Silhouette coefficient and qualitatively discussed. The proposed method was also compared to an implementation based on the standard density-based spatial clustering of applications with noise algorithm, where it showed more promising results.</p>}},
author = {{Farnaghi, Mahdi and Ghaemi, Zeinab and Mansourian, Ali}},
issn = {{2095-0055}},
keywords = {{Disaster management; Hurricane Florence; Natural language processing; Spatio-temporal tweet analysis; Tweet clustering; Twitter; Machine Learning (ML)}},
language = {{eng}},
number = {{3}},
pages = {{378--393}},
publisher = {{Springer}},
series = {{International Journal of Disaster Risk Science}},
title = {{Dynamic Spatio-Temporal Tweet Mining for Event Detection : A Case Study of Hurricane Florence}},
url = {{http://dx.doi.org/10.1007/s13753-020-00280-z}},
doi = {{10.1007/s13753-020-00280-z}},
volume = {{11}},
year = {{2020}},
}