Identifying Crisis Response Communities in Online Social Networks for Compound Disasters The Case of Hurricane Laura and Covid -19 Khondhaker Al Momin

2025-05-08 0 0 2.34MB 28 页 10玖币

侵权投诉

Identifying Crisis Response Communities in Online Social Networks for

Compound Disasters: The Case of Hurricane Laura and Covid-19

Khondhaker Al Momin

Graduate Research Assistant

School of Civil Engineering & Environmental Science

University of Oklahoma

202 W. Boyd St., Norman, OK 73019-1024

Email: momin@ou.edu

H M Imran Kays

Graduate Research Assistant

School of Civil Engineering & Environmental Science

University of Oklahoma

202 W. Boyd St., Norman, OK 73019-1024

Email: kays@ou.edu

Arif Mohaimin Sadri, Ph.D.

Assistant Professor

School of Civil Engineering & Environmental Science

University of Oklahoma

202 W. Boyd St., Norman, OK 73019-1024

Email: sadri@ou.edu

(Corresponding Author)

Momin et al.

ABSTRACT

Online social networks allow different agencies and the public to interact and share the underlying risks

and protective actions during major disasters. This study revealed such crisis communication patterns

during hurricane Laura compounded by the COVID-19 pandemic. Laura was one of the strongest (Category

4) hurricanes on record to make landfall in Cameron, Louisiana. Using the Application Programming

Interface (API), this study utilizes large-scale social media data obtained from Twitter through the recently

released academic track that provides complete and unbiased observations. The data captured publicly

available tweets shared by active Twitter users from the vulnerable areas threatened by Laura. Online social

networks were based on Twitter’s user influence feature (i.e., mentions or tags) that allows notifying other

users while posting a tweet. Using network science theories and advanced community detection algorithms,

the study split these networks into twenty-one components of various size, the largest of which contained

eight well-defined communities. Several natural language processing techniques (i.e., word clouds,

bigrams, topic modeling) were applied to the tweets shared by the users in these communities to observe

their risk-taking or risk-averse behavior during a major compounding crisis. Social media accounts of local

news media, radio, universities, and popular sports pages were among those who involved heavily and

interacted closely with local residents. In contrast, emergency management and planning units in the area

engaged less with the public. The findings of this study provide novel insights into the design of efficient

social media communication guidelines to respond better in future disasters.

Keywords: Hurricane, Crisis Respond, First Responder, Social Network, Twitter Data

Momin et al.

INTRODUCTION

Natural, man-made, and technological disasters have major adverse consequences for society and

the economy. Affected communities must be well-informed about the potential distress they may encounter

during a disaster in order to take appropriate preparation mentally and logistically. Emergency management

authorities must be aware of the needs and concerns of affected people to respond effectively throughout a

crisis (1). Understanding the communication pattern, perspectives, thoughts, and needs of the affected

people, as well as proper information dissemination, is pivotal during any disaster or emergency situation.

On August 27, 2020, hurricane Laura, one of the most powerful and deadliest “category 4”

hurricane made landfall with peak intensity in Cameron, Louisiana (2) in the middle of the Covid-19

pandemic. An estimated 20 million people were in the path of the storm, with 1.5 million in Texas and

Louisiana being ordered to evacuate (3). Hurricane Laura spawned a 15-feet storm surge that drenched

some regions with 10 inches of rain (ref. Figure 1) and spawned four tornadoes (4). Lesser Antilles, Greater

Antilles, Bahamas, Gulf Coast of the United States (U.S.), the Midwestern and Eastern U.S. have all been

heavily affected by hurricane Laura. It largely impacted Louisiana, and a total of 14 persons were killed

(10 deaths in Louisiana and 4 in Texas) in the U.S., 31 in Haiti, and 3 in the Dominican Republic (5-7).

More than 750,000 people experienced a significant power outage due to hurricane Laura (8), while many

refineries in the U.S. were shut down for several weeks pending repairs and power restoration in the

aftermath of hurricane Laura (9). Moreover, this “category 4” hurricane struck the U.S. during a pandemic,

when people were instructed to keep their distance. Such crisis further escalates when there is a

communication gap between vulnerable communities and emergency management agencies (10).

Figure 1: Heavy Rainfall due to Hurricane Laura

According to the U.S. Census Bureau, 84 percent of U.S. households own a cell phone, and 78

percent own a desktop or laptop computer (11). During a disaster, people can communicate in person or via

mobile phone. However, by using online social media platforms (SMPs), they can reach and connect with

more people in a much shorter period of time, which is why the use of online social media during disasters

has recently increased (12). Many studies have shown that weather and situational awareness information

is widely disseminated through SMPs (i.e., Facebook, Twitter, Reddit, Instagram, etc.) (13-17). These

SMPs allow users to share ideas, thoughts, and information through virtual networks in a timely manner

Momin et al.

(18). Users readily share real-time information on social media about a variety of topics such as disasters,

potential crises, weather updates, and traffic updates, among others. According to a recent report (19), there

are approximately 450 million monthly active Twitter users (20), and among them, more than 80.9 million

users are from the U.S. The majority of the emergency management and law enforcement agencies in the

U.S. have a Twitter account where they share important and relevant information (21). People follow

different organizations as well as different celebrities on social media, and a lot of users are directly

influenced by their followees on SMPs (22).

The Covid-19 outbreak saw more than 29.7% more users spending 1-2 hours per day on social

media, with 20.5% using SMPs 30 minutes to an hour more (18,23), and even during a power outage, people

can still use different social media to keep communication with their friends and family (16). This makes

social media a viable data source for user-generated content, particularly during a disaster when traditional

surveys are inconvenient. Many literatures have demonstrated the effectiveness of social media data in

disaster management, particularly in crisis communication (16,24-29), human mobility (30-33), damage

valuation (34), and event detection (35,36), among others. However, there are very few studies that have

looked at the social network connectivity of the local people and network-level properties of the local

communities along with their crisis narrative analysis during a compound hazard like hurricane Laura amid

the Covid 19 pandemic.

Twitter recently launched its Academic Application Programming Interface (API) (37), which

provides a full history of public conversation through full-archive search endpoint (38), thus making Twitter

a reliable social media data source. In this study, a user-mention network has been developed from the local

Twitter user who tweeted during hurricane Laura and tweeted from the affected areas. The study is

concerned with the following three research questions:

(i) Who are the agents of the social network, and what are their communication patterns?

(ii) How does crisis communication differ from one community to another community?

(iii) What is the role of different agencies in the social network?

The findings of this study provide novel insights on how to create effective social media

communication guidelines so that future disasters can be handled more effectively. The communication

pattern in the local community is observed in this study with the local connectivity of different agents.

RELATED WORK

According to global data portal by Kepois, the emergence of SMPs has been met by a spike in

social media users of an excess of 4.70 billion people as of 2022, equating to ~59% of the total global

population (18). People are becoming accustomed to relying more on social media than on mainstream

news sources (39). SMPs act as a global connecting point for immediate response and news dissemination;

for example, according to a recent study, Twitter alone generates more than 143K trending tweets per

second (40). These SMPs are extremely effective communication tools, particularly during emergencies

such as hurricanes, and they serve as a hub for public opinion and emotional guidance during crisis

situations (41). Many researchers have used social media data in analyzing a variety of natural and man-

made hazards, including fire (42), flood (43), tsunami (44), earthquake (44,45), hurricanes (46,47), and

active shooting (39,48) among others.

Social media play a great role during an emergency like hurricane evacuation. The routing decision,

evacuation destination, and evacuation decisions can be influenced by social media (49). Social media, as

a new data source, can provide real-time information for flood monitoring (43). Extreme weather

consequences can be seen in real-time on social media (50). Mohanty et al. found that the hurricane's per-

capita economic impact is significantly correlated with per-capita Twitter engagement (27). In the classical

Momin et al.

evacuation literature, there have been a lot of studies on disaster response, particularly on hurricane

evacuation (51-59) and wildfire (60-62), using traditional survey data, but according to recent studies, social

media plays an important role in catastrophes by providing a participative, collaborative, and self-organized

structure for communicating and disseminating situational information (25,63,64). The social network also

has a great influence on decision-making, and decision-makers follow their peers during similar events

(65). Topic modeling, specifically Latent Dirichlet Allocation (LDA), is used in disaster-related

information retrieval from social media data. Several studies used Twitter data to identify disaster-related

topics using a static (40,66-68) and dynamic (69) LDA model. Yang et al. proposed a method for

incorporating sentiment with a topic model (70); however, none of these studies consider any social network

properties. Rajput et al. have explored the network properties of the different local agencies before, during,

and after Hurricane Harvey (71). Sadri et al. found that users at the network's core are less eccentric and

have higher degrees, and they are more engaged in disseminating information (47). It is pivotal to study

how people interact in online social networks in order to develop and implement more effective guidelines

in times of crisis (67,72).

However, very few studies have explored the connectivity of local people on social media during

a compound crisis (a hurricane in the midst of a pandemic) as well as their communication patterns from a

social network perspective. This study has explored the communication patterns of the locally affected

people and connectivity with themselves and with different types of agencies from a user-mention network.

DATA DESCRIPTION AND PREPROCESSING

There were several hurricanes in the year 2020. Hurricane Laura was the worst of them for the

United States, killing at least ten people in Louisiana (73). On August 20, 2020, a tropical depression formed

in the Cape Verde islands in the central Atlantic Ocean (74) and was officially named “Laura” on August

21, 2020 (75), which eventually became a category four hurricane and made landfall in Cameron, Louisiana,

USA on August 27, 2020 (76). People freely express their opinion on social media like Twitter. Following

is one of the tweets generated from Louisiana during the hurricane Laura: “Me waiting patiently to get the

internet and cable fix and back on.... I hate hurricane season”. One of the news channels tweeted regarding

the devastation in the following tweet: “@fox4beaumont Hurricane Laura death toll in Louisiana rises to

10”. In this study, Twitter data was used to identify crisis response communities in online social networks.

The Twitter Academic Application Programming Interface (API) (37), which provides a full history

of public conversation through full-archive search endpoint (38), was used to get Twitter data from August

13, 2020, to September 03, 2020. Different query options have been used to make sure that all location-

based data is collected. At first, ‘(Hurricane OR Hurricane Laura) has:geo’ query option was used to collect

all of the tweets regarding Hurricane Laura. Secondly, the “place: ‘Louisiana’” query option was used to

collect all of the generated tweets tagged with Louisiana within the mentioned time frame. Finally, the

"point radius" query option with a 25-mile radius circle was used to ensure that all location-based tweets

were collected from the highly impacted locations in Louisiana (Figure 2). The API sets several

constraints for geolocation-based data collection, i.e., the maximum radius of the "point radius"

cannot be greater than 25 miles (38). To overcome this constraint, multiple "point radius" queries

were used on these highly affected locations.

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

IdentifyingCrisisResponseCommunitiesinOnlineSocialNetworksforCompoundDisasters:TheCaseofHurricaneLauraandCovid-19KhondhakerAlMominGraduateResearchAssistantSchoolofCivilEngineering&EnvironmentalScienceUniversityofOklahoma202W.BoydSt.,Norman,OK73019-1024Email:momin@ou.eduHMImranKaysGraduateResearchAss...

展开>> 收起<<

Identifying Crisis Response Communities in Online Social Networks for Compound Disasters The Case of Hurricane Laura and Covid -19 Khondhaker Al Momin.pdf

共28页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Identifying Crisis Response Communities in Online Social Networks for Compound Disasters The Case of Hurricane Laura and Covid -19 Khondhaker Al Momin

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: