Encyclopedia
Please note this is an old version of this entry, which may differ significantly from the current revision.
Subjects:
Behavioral Sciences
During the COVID-19 pandemic, social media has become an emerging platform for the public to find information, share opinions, and seek coping strategies. Vaccination, one of the most effective public health interventions to control the COVID-19 pandemic, has become the focus of public online discussions. Several studies have demonstrated that social bots actively involved in topic discussions on social media and expressed their sentiments and emotions, which affected human users. However, it is unclear whether social bots’ sentiments affect human users’ sentiments of COVID-19 vaccines.
- social bots
- sentimental engagement
- COVID-19
1. Introduction
Social media has become a major mode of global communication, characterized by its potential to reach large audiences and spread information rapidly [1,2]. In contrast to traditional media, content posted need not undergo editorial management nor scientific vetting, and users frequently maintain anonymity, which allow individuals to express their views directly [3,4]. This means that there are a multitude of emotional expressions, false information, and rumors on social media, which could deviate the public opinions from the track of reason [5]. Researchers thus argue that it is important to analyze sentiments on social media [6,7,8].
In face of health emergencies, social media has become an emerging platform for the public to share opinions, express attitudes, and seek coping strategies [9,10,11,12]. During the outbreak of coronavirus disease 2019 (COVID-19), intensive global efforts toward physical distancing and isolation to curb the spread of COVID-19 may have intensified the use of social media for individuals to remain connected [13]. Researchers found that the public express their fears of infection and shock regarding the contagiousness of COVID-19 [14], along with their feelings about infection control strategies on social media [15,16,17]. However, established evidence has proven that social bots—automated accounts controlled and manipulated by computer algorithms [18]—often express sentiments in online discussions on social media [7,19], which might manipulate or even distort online public opinions [20,21].
The birth of COVID-19 vaccines has also sparked heated discussions and debates on social media [22]. Social media has long been a prominent sphere of vaccine debates [23]. With 217 million daily activities [24], Twitter is a convenient tool for discussing and debating public health topics, including vaccines and vaccination [2]. Both pro-vaccine and anti-vaccine information is prevalent on Twitter [2]. Social bots disseminate anti-vaccine messages by masquerading as legitimate users, eroding the public consensus on vaccination [25].
A previous study found that social bots have three high-intensity stages of intervention in the COVID-19 vaccines-related topics [22]. Yet, how social bots shape public sentiment and public opinion remains to be studied.
Against this backdrop, this study seeks to scrutinize whether the sentiments of social bots affect human users’ sentiments of COVID-19 vaccines. This overall objective is divided into the following two sub-objectives.
The first sub-objective is to analyze the sentiments of both social bots and human users on the discussion of COVID-19 vaccines in three stages. The semi-supervised deep learning model, i.e., the BERT-CNN sentiment analysis framework, was used for the sentiment analysis of tweets posted by social bots and humans in this study. The second sub-objective is to perform a statistical analysis of whether there was a time-series correlation between the sentiments of social bots and humans.
2. Vaccine Discussions and Sentiments on Social Media
Vaccine discussions are widespread on social media. The discoursing trends of vaccines on social media are linked to real-world events [26]. Gunaratne et al. found that anti-vaccination discourses on Twitter soared in 2015, coinciding with the 2014–2015 measles outbreak. After the COVID-19 outbreak, social media became one of the main spheres for the debates and discussions related to COVID-19 and its vaccines [13]. Especially, discussions of COVID-19 vaccines are closely related to real-world events [22]. For example, as the Delta virus ravaged the world in June–July 2021, public discussion about vaccines on social media platforms mainly revolved around whether the COVID-19 vaccines could prevent the Delta variant [22].
People tend to express their sentiments in online vaccine discussions [13]. Generally speaking, one sentiment is to support vaccines and vaccination, and the other is against them. The topics of pro-vaccine and anti-vaccine have been developed on social media and have received the attention of researchers. Kang et al. examined current vaccine sentiments on social media by constructing and analyzing semantic networks of vaccine information from highly shared websites of Twitter users in the United States. The semantic network of positive vaccine sentiments demonstrates greater cohesiveness in discourse compared to the larger, less-connected network of negative vaccine sentiments [6]. Melton et al. analyzed the COVID-19 vaccine-related content and found that sentiments expressed in these social media communities are more positive but have had no meaningful change since December 2020 [27].
Misinformation [8], individual opinions [23], and health literacy [28] affect public sentiments of vaccination. The prevalence of negative vaccine sentiments is generally skeptical and distrustful of government organizations [6]. Public sentiments related to vaccination may be affected by technical factors. For example, the information delivered by a chatbot that answers people’s questions about COVID-19 vaccines positively affects attitudes and intentions toward COVID-19 vaccines [29].
3. Sentiment Engagement of Social Bots in Online Vaccine Discussions
Since social bots under algorithmic control are able to send messages continuously and frequently [18], they could even guide or influence public sentiments in social media [30]. In the process of interacting with humans, bots in a social system affect the way humans perceive social norms and increase human exposure to polarized views. For example, increasing their exposure to harmful contents and inflammatory narratives by disseminating large amounts of contents polarizes humans’ sentiments and emotions [31]. With the intervention of social bots, the discussion around a specific topic can easily be separated from the event and become a battle between two extreme emotions, which affect the public’s emotions and attitudes [20].
Since 2010, social bots have been active in political elections, business events, health communication, and other fields, presenting a strong influence on public sentiments [32,33]. In health-related topics, social bots have been accused of spreading misinformation, promoting polarized opinions, and manipulating public sentiments [7]. For example, social bots can influence public attitudes about the efficiency of cannabis in dealing with mental and physical problems [21]. During the COVID-19 pandemic, social bots spread and amplify false medical information and conspiracy theories, which influence the public’s correct judgments and stimulate their negative emotions [34].
Social bots have been active in various vaccine topics and guiding public sentiments for a long time. The messages spread by some social bots were more divisive and political, and their disseminations of anti-vaccine contents weakened the public consensus on vaccination and shook public confidence in vaccination [25]. Egli’s research on anti-vaccine social bots suggests that bots can arouse people’s resistance to COVID-19 vaccines and promote the spread of conspiracy theories by spreading carefully fabricated vaccine rumors [35].
4. Social Media and Sentiment Analysis
Sentiment analysis is a computer process developed to categorize users’ opinions or sentiments expressed in source texts [36]. It involves categorizing subjective opinions from text, audio, and video sources to determine polarities (e.g., positivity, negativity, and neutrality), emotions (e.g., anger, sadness, and happiness), or states of mind toward target topics, themes, or aspects of interest [37]. Currently, automated sentiment analysis receives increasing attention from the academia [38] and has become one of the key techniques for processing large amounts of social media data [39]. In methodology, artificial intelligence (AI) techniques such as machine learning (ML), deep learning (DL), and natural language processing (NLP) can extract topics and sentiments from largely unstructured social media data [37].
Researchers have used automated sentiment analysis to analyze the sentiments and emotions of social bots [7,40]. Yuan et al. combined sentiment analysis with supervised machine learning (SML) and community detection on social networks to unveil the communication patterns of pro-vaccine and anti-vaccine social bots on Twitter [40]. Linguistic inquiry and word count (or LIWC) was used to analyzed samples of social bots and individual accounts to investigate the different sentiment and emotion components in COVID-19 pandemic discussions on Twitter [7].
According to above references, social bots actively involved in topic discussions on social media can express their sentiments [7,20], which affect human users [35]. Social bots also participate in COVID-19 vaccine discussions on Twitter [22]. However, it is unclear on how the sentiments of social bots are expressed in online discussions about COVID-19 vaccines and whether their sentiments affect human users’ sentiments of COVID-19 vaccines. Based on previous study, the work used semi-supervised deep learning to continue the sentiments analysis of tweets posted by social bots and human users. Then, the Granger causality test was used to analyze whether there was a time-series causality between the sentiments of social bots and humans.
This entry is adapted from the peer-reviewed paper 10.3390/su14095566
This entry is offline, you can click here to edit this entry!
