2019
DOI: 10.1002/asi.24208
|View full text |Cite
|
Sign up to set email alerts
|

Early detection of heterogeneous disaster events using social media

Abstract: This article addresses the problem of detecting crisisrelated messages on social media, in order to improve the situational awareness of emergency services. Previous work focused on developing machine-learning classifiers restricted to specific disasters, such as storms or wildfires. We investigate for the first time methods to detect such messages where the type of the crisis is not known in advance, that is, the data are highly heterogeneous. Data heterogeneity causes significant difficulties for learning al… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
15
0

Year Published

2019
2019
2023
2023

Publication Types

Select...
6
3
1

Relationship

0
10

Authors

Journals

citations
Cited by 29 publications
(15 citation statements)
references
References 18 publications
0
15
0
Order By: Relevance
“…Their event detection probability threshold was 0.95. On the other hand, Pekar et al [36] proposed an approach that can detect real-world disaster events from social media data without any prior knowledge about the events. They tackled this issue by implementing a new ensemble method that utilized the available data on the training dataset.…”
Section: Shallow-ml-based Approachesmentioning
confidence: 99%
“…Their event detection probability threshold was 0.95. On the other hand, Pekar et al [36] proposed an approach that can detect real-world disaster events from social media data without any prior knowledge about the events. They tackled this issue by implementing a new ensemble method that utilized the available data on the training dataset.…”
Section: Shallow-ml-based Approachesmentioning
confidence: 99%
“…Various methods exist to differentiate between disaster types within a dataset, however, most consist of an ad hoc review for type-specific terms and phrases within an event report [63,64]. This highlights the importance of identifying terms and phrases that are representative of the subtype of interest [65].…”
Section: An Index For Flood Type Classificationmentioning
confidence: 99%
“…These social media data with temporal and spatial attributes have become an important means of understanding public behavior [ 30 ]. Managers and researchers can analyze social media data for disaster detection [ 31 , 32 ], risk communication [ 12 , 33 , 34 , 35 , 36 ], intelligent decision-making [ 37 ], and emergency response [ 38 , 39 ]. For example, government social media can be used for increasing vigilance and awareness in the prodromal stage, disseminating information and increasing transparency in the acute stage, and focusing on mental health support and recovery policies in the chronic stage [ 40 ].…”
Section: Literature Reviewmentioning
confidence: 99%