Social Media Messages Research Articles

Analyzing the Effect of Social Media Sentiment on Stock Prices

As consumers increasingly use social media to find out about a company and its products, the sentiment of that social media become even more important. This study is exploratory in nature, the aim is to determine if social media sentiment influences the stock price of large global companies. Stock prices and annual revenue was measured for Dow 30 stocks at two time periods six months apart. Social Media sentiment is more difficult to measure. In this study the sentiment ratings from Social Searcher.com were used to indicate whether social media messages by the public are positive, negative or neutral. While past studies have focused on either Facebook or Twitter messages, this tool aggregates messages from eleven social media platforms globally. In conducting this study numerous limitations and recommendations for future studies became apparent. The value of this exploratory study is in the flaws exposed by conducting this study. This provides a rich template for future studies on social media sentiment and its effect on a company.

Rumor Classification through a Multimodal Fusion Framework and Ensemble Learning.

The proliferation of rumors on social media has become a major concern due to its ability to create a devastating impact. Manually assessing the veracity of social media messages is a very time-consuming task that can be much helped by machine learning. Most message veracity verification methods only exploit textual contents and metadata. Very few take both textual and visual contents, and more particularly images, into account. Moreover, prior works have used many classical machine learning models to detect rumors. However, although recent studies have proven the effectiveness of ensemble machine learning approaches, such models have seldom been applied. Thus, in this paper, we propose a set of advanced image features that are inspired from the field of image quality assessment, and introduce the Multimodal fusiON framework to assess message veracIty in social neTwORks (MONITOR), which exploits all message features by exploring various machine learning models. Moreover, we demonstrate the effectiveness of ensemble learning algorithms for rumor detection by using five metalearning models. Eventually, we conduct extensive experiments on two real-world datasets. Results show that MONITOR outperforms state-of-the-art machine learning baselines and that ensemble models significantly increase MONITOR’s performance.

Local Governments’ Framing of Social Media Messages and Citizens’ Responses (WITHDRAWN)

Public administrations frequently communicate with citizens and often use social media to do so. The importance of such communication is elevated in times of crisis. Citizens have an increased demand for information and interaction and administrations want to limit the extent and consequences of the crisis. Previous research has analyzed how public administrations communicate during a crisis, but little is known about how citizens respond to these messages and how their responses change based on the framing of administrations’ messages. We analyze how different rhetorical frames used in local governments’ social media messages affect three types of citizen responses: approval, diffusion, and engagement. We hypothesize several relationships based on the expectancy-disconfirmation framework and emotions as social information theory. The hypothesized effects are tested on a novel dataset of German local governments’ Twitter communications consisting of more than 50,000 messages posted over a period of six months and more than 700,000 reactions to these messages. The results show that the use of analytical framing increases citizens’ approval, diffusion, and engagement. Affective framing is found to positively influence approval, while negatively affecting diffusion and engagement. Disentangling what drives these diverging effects, we show that positive emotional framing positively affects citizens’ approval and negatively affects diffusion and engagement. On the other hand, negative emotions increase citizens’ diffusion and engagement. Particularly these contrasting effects emphasize the need to better understand the role of rhetorical framing in public sector communication, as rhetorical frames might affect citizen reactions in opposing ways. Besides these theoretical implications, several managerial implications for public sector communication can be derived.

AI-SocialDisaster: An AI-based software for identifying and analyzing natural disasters from social media

AI-SocialDisaster is a decision support system for identifying and analyzing natural disasters like earthquakes, floods, bushfires using social media feeds. It captures real-time social media messages and then uses Natural Language Processing (NLP) based algorithms like entity detection, category classification, and sentiment analysis to identify and locate various natural disasters. Moreover, using Artificial Intelligence (AI) based algorithms like anomaly detection, regression, and clustering, AI-SocialDisaster generates AI-based insights for disaster planners and strategists. The software can be accessed through Windows, iOS, and Android apps from a wide range of devices including mobiles, tablets, and desktops. AI-SocialDisaster is available at https://github.com/DrSufi/DisasterAI.

Relationship of Social Media and Body Image Dissatisfaction among University Students

Social media has become an integral part of young adults’ lives today. It has moved well beyond simple entertainment and now can profoundly affect many areas of functioning. Social media messages promote the ideal, attractive body images that affect the university student’s attitudes toward the body. The current study examined the relationship between social media and body image dissatisfaction and examined how social media impacts body image among university students. Sample size estimation was calculated from G*power and 462 students minimum required in the current study. This cross-sectional study recruited (N=500) respondents from different universities such as the University of Sahiwal, Government college university of Faisalabad Sahiwal campus, Virtual University of Sahiwal, University of Okara, Virtual University of Okara, University of Management and Technology, and the University of Lahore by employing simple random sampling. There were two questionnaires applied the social media questionnaire and Socio-Cultural Attitudes towards Appearance Scale-3 (SATAQ-3) for data collection. Data were analyzed with SPSS version 25.0 and results showed that there was a significant negative relationship of social media with body image dissatisfaction. Moreover, male students were higher users of social media as compared to female students as well as higher body image dissatisfaction as compared to girls. Implications and limitations of the study are mentioned as well as future directions also suggested

Adaptation and Dissemination of a National Cancer Institute HPV Vaccine Evidence-Based Cancer Control Program to the Social Media Messaging Environment.

Social media offers a unique opportunity to widely disseminate HPV vaccine messaging to reach youth and parents, given the information channel has become mainstream with 330 million monthly users in the United States and 4.2 billion users worldwide. Yet, a gap remains on how to adapt evidence-based vaccine interventions for the in vivo competitive social media messaging environment and what strategies to employ to make vaccine messages go viral. Push-pull and RE-AIM dissemination frameworks guided our adaptation of a National Cancer Institute video-based HPV vaccine cancer control program, the HPV Vaccine Decision Narratives, for the social media environment. We also aimed to understand how dissemination might differ across three platforms, namely Instagram, TikTok, and Twitter, to increase reach and engagement. Centering theory and a question-answer framework guided the adaptation process of segmenting vaccine decision story videos into shorter coherent segments for social media. Twelve strategies were implemented over 4 months to build a following and disseminate the intervention. The evaluation showed that all platforms increased following, but Instagram and TikTok outperformed Twitter on impressions, followers, engagement, and reach metrics. Although TikTok increased reach the most (unique accounts that viewed content), Instagram increased followers, engagement, and impressions the most. For Instagram, the top performer, six of 12 strategies contributed to increasing reach, including the use of videos, more than 11 hashtags, COVID-19 hashtags, mentions, and follow-for-follow strategies. This observational social media study identified dissemination strategies that significantly increased the reach of vaccine messages in a real-world competitive social media messaging environment. Engagement presented greater challenges. Results inform the planning and adaptation considerations necessary for transforming public health HPV vaccine interventions for social media environments, with unique considerations depending on the platform.

Social Media Messaging, Fake News, and COVID-19 outbreak in Lagos, Nigeria

COVID-19 is a public health concern that has negatively affected human development across the globe. This study examined the attitude towards social media messaging on fake news about COVID-19 outbreak in Lagos, Nigeria. This study was hinged on the uses and gratification theory to understand how individuals utilize social media to publish fake news reports on COVID-19. A mixed method approach (survey and ethnographic observation) was adopted to achieve the objectives of the study. The survey method served as a primary method to generalize inferences and also determine individual perception on social media messaging and fake news reports, while ethnographic observation served as the secondary method to generate rich descriptive data that captures the experiences, feelings and behaviors of individuals. In addition, the Spearman correlation analysis and Somer D’ measurement were used to test a hypothesis to ascertain the possibility of connection between variables. The results indicated that a significant majority of the respondents were sufficiently aware of social media reports on COVID-19 outbreak in Lagos State. In addition, the findings show that 83.1% of the respondents were adequately aware of fake news reports on social media about COVID-19 outbreak. Furthermore, the results reveal that almost 67% of the respondents had cautioned individuals about the circulation of fake news report about COVID-19 on social media platforms. It was recommended that governments, international agencies and non-governmental agencies should partner to develop a media literacy project on COVID-19 to educate and sensitive individuals about how to detect fake news as well as the importance of priority health practices such as vaccination, healthy lifestyle practices to curb the threat of COVID-19 in Nigeria and other parts of the world.

Exploring Donor Influence and Public Engagement: Computational and Thematic Analyses of Social Media Messages

Exploring Donor Influence and Public Engagement: Computational and Thematic Analyses of Social Media Messages

Effectiveness of Mediterranean Diet on Diabetic Control and Cardiovascular Risk Modification Among Patients With Type 2 Diabetes Mellitus in Oman: A Study Protocol

Introduction: In the recent years, there is remarkable increased in the prevalence of Type 2 Diabetes Mellitus (T2DM) in Middle East countries including Oman. There is good evidence that Mediterranean Diet (MedDiet) is effective over diabetes control and several cardiovascular risk factors in different populations, with little compelling evidence among Omanis. This paper describes the protocol of a wait-list, open labelled, randomized control trial, with its main objective aimed to determine the effectiveness of Mediterranean Diet intervention on glycaemic control and cardiovascular risks among T2DM patients in Oman. Methods: A total of 140 eligible T2DM patients will be recruited. Participants in the intervention group will undergo a six-month MedDiet program comprising of different activities (individual dietary counseling, cooking classes, phone calls and social media messages) while the control group will continue with standard diabetes care. Data collection will be conducted at baseline, after three and six months. The 2 x 3 mixed-design ANOVA will be used to determine the mean changes in outcome variables over the full study period between the two groups. Discussion: Epidemiology studies on nutrition and health had focused on dietary pattern, which provides an opportunity to account for nutrient-nutrient interactions lately. Mediterranean Diet has produced consistent findings on its protective role in diabetes management, with little information on its effectiveness in population outside of Mediterranean basin, including Oman. The outcomes of current study will be used to inform community and health care professionals on the effectiveness and practically of MedDiet on diabetes management.

Strategi Pesan Persuasif di Media Sosial Instagram @somethincofficial

Social media is a tool used for various activities, including supporting business and marketing. Small and medium industries tend to choose to use Instagram to communicate with young and productive consumers, the majority of whom are similar to the characteristics of these social media users. SOMETHINC is a local skincare and beauty brand made in Indonesia that actively uses Instagram to interact with its consumers. This study aims to analyze how communication messages in SOMETHINC's Instagram social media can affect consumers' cognitive processes in making choices to use products based on central and peripheral concepts in the Elaboration Likelihood Model theory. Based on the Elaboration Likelihood Model developed by Petty and Cacioppo (1983), the central pathway is a pathway that encourages individual behavior change after going through a process of careful and wise consideration of the benefits of information provided as an advocacy effort. Meanwhile, the peripheral pathway explains that behavior change occurs without going through a process of careful consideration of the benefits of the information conveyed and is more influenced by some exciting signs, cues, or sources in persuasion efforts to encourage behavior change. This research method uses a digital ethnographic method with the primary data source in the form of documentation of the Instagram account @somethincofficial. The results show that SOMETHINC has used strategies according to the four pillars of social media and persuasive communication messages according to the central and peripheral pathways in the Elaboration Likelihood Model. This study concludes that the Elaboration Likelihood Model approach can help brands design more effective social media strategies to adapt persuasive messages to the intended audience.

The ‘Fauci Effect’: Reducing COVID-19 misconceptions and vaccine hesitancy using an authentic multimodal intervention

Social media environments enable users to proliferate misinformation surrounding COVID-19. Expert sources, such as Dr. Anthony Fauci have leveraged social media to present corrective multimedia messages. However, little is known about the efficacy of these messages in revising common misconceptions about COVID-19 and influencing behavior. In this study, we examined the efficacy of a multimodal intervention using authentic social media messages that directly addressed common COVID-19 misconceptions. Going further, we identified individual differences that influenced the effectiveness of the intervention, as well as whether those factors predicted individuals’ willingness to receive a COVID-19 vaccine. The results showed that the intervention was successful in increasing knowledge when compared to a baseline control. Those who were older and reported less vaccine hesitancy showed greater learning from the intervention. Factors that significantly predicted intention to vaccinate included receiving the intervention, vaccine confidence, vaccine hesitancy, prior flu vaccination history, age, and fear of COVID-19. These findings indicate that multimodal messages can be effectively leveraged in social media to both fight misinformation and increase intention to be vaccinated - however, these interventions may not be as effective for all individuals.

Extracting locations from sport and exercise-related social media messages using a neural network-based bilingual toponym recognition model

Sport and exercise contribute to health and well-being in cities. While previous research has mainly focused on activities at specific locations such as sport facilities, "informal sport" that occur at arbitrary locations across the city have been largely neglected. Such activities are more challenging to observe, but this challenge may be addressed using data collected from social media platforms, because social media users regularly generate content related to sports and exercise at given locations. This allows studying all sport, including those "informal sport" which are at arbitrary locations, to better understand sports and exercise-related activities in cities. However, user-generated geographical information available on social media platforms is becoming scarcer and coarser. This places increased emphasis on extracting location information from free-form text content on social media, which is complicated by multilingualism and informal language. To support this effort, this article presents an end-to-end deep learning-based bilingual toponym recognition model for extracting location information from social media content related to sports and exercise. We show that our approach outperforms five state-of-the-art deep learning and machine learning models. We further demonstrate how our model can be deployed in a geoparsing framework to support city planners in promoting healthy and active lifestyles.

Collection of Road Traffic Incidents in Bangkok from Twitter Data based on Deep Learning Algorithm

Text processing technology from Twitter to report notification formats that are known in many countries with verification on different languages. This research presents the development of a neural network memory learning model. To solve the problem of classifying incidence patterns and identifying severity of incidents from Thai social media messages. For gathering incident data and reporting incidents externally from a single reporting platform by using deep learning models like MLP, CNN and LSTM which is designed by dividing the study into 3 types, including examination traffic incidence identification pattern that can identify the report as general news or traffic reporting Incident Identification Patterns. These include traffic conditions, accidents, disasters, damaged roads, or other than the aforementioned patterns, and the pattern indicating the severity of the incidence consists of normal level, medium level and lane blocking or stationary levels. The results demonstrated the ability of LSTM learning with the best results in incidence detection and incidence pattern identification at 93.44% and 87.40%, respectively, and the CNN method was able to State the severity of the incidence at best, reaching 91.42%.

Collection of Road Traffic Incidents in Bangkok from Twitter Data based on Deep Learning Algorithm

Text processing technology from Twitter to report notification formats that are known in many countries with verification on different languages. This research presents the development of a neural network memory learning model. To solve the problem of classifying incidence patterns and identifying severity of incidents from Thai social media messages. For gathering incident data and reporting incidents externally from a single reporting platform by using deep learning models like MLP, CNN and LSTM which is designed by dividing the study into 3 types, including examination traffic incidence identification pattern that can identify the report as general news or traffic reporting Incident Identification Patterns. These include traffic conditions, accidents, disasters, damaged roads, or other than the aforementioned patterns, and the pattern indicating the severity of the incidence consists of normal level, medium level and lane blocking or stationary levels. The results demonstrated the ability of LSTM learning with the best results in incidence detection and incidence pattern identification at 93.44% and 87.40%, respectively, and the CNN method was able to State the severity of the incidence at best, reaching 91.42%.

COVID-19 vaccine hesitancy: a social media analysis using deep learning.

Hesitant attitudes have been a significant issue since the development of the first vaccines—the WHO sees them as one of the most critical global health threats. The increasing use of social media to spread questionable information about vaccination strongly impacts the population’s decision to get vaccinated. Developing text classification methods that can identify hesitant messages on social media could be useful for health campaigns in their efforts to address negative influences from social media platforms and provide reliable information to support their strategies against hesitant-vaccination sentiments. This study aims to evaluate the performance of different machine learning models and deep learning methods in identifying vaccine-hesitant tweets that are being published during the COVID-19 pandemic. Our concluding remarks are that Long Short-Term Memory and Recurrent Neural Network models have outperformed traditional machine learning models on detecting vaccine-hesitant messages in social media, with an accuracy rate of 86% against 83%.

HPV and COVID-19 vaccines: Social media use, confidence, and intentions among parents living in different community types in the United States.

Our study measured parental confidence and intention/uptake of two adolescent vaccines (HPV and COVID-19), focusing on differences among community types including urban, suburban, and rural. Although social media provides a way for misinformation to spread, it remains a viable forum for countering misinformation and engaging parents with positive vaccine information across community types. Yet, little is understood about differences in social media use and vaccine attitudes and behaviors for parents living in rural, suburban and urban areas. We sought to determine how to better reach parents living in different community types with targeted social media channels and messaging. In August 2021, we used a cross-sectional survey programmed in Qualtrics to collect data from 452 parents of children ages 9 to 14 living in different community types across the United States. Participants came from a survey panel maintained by CloudResearch. Survey questions asked about demographics, political affiliation, community type, social media use, health and vaccine information sources, and attitudes and behaviors regarding the HPV and COVID-19 vaccines. Our sample of parents (n = 452) most frequently used Facebook (76%), followed by YouTube (55%), and Instagram (43%). When comparing social media use by community type, parents used the top platforms at similar rates. Social media use was associated with vaccine confidence and intention/uptake in unadjusted models but not in adjusted models. Further, there were no significant differences in HPV vaccine confidence or intention/uptake by community type (i.e., rural, suburban, urban). For the COVID-19 vaccine, parents in rural communities were less likely to have vaccine confidence and intention/uptake in the unadjusted model. For both HPV and COVID-19 vaccines, political affiliation was the only common factor associated with both vaccine confidence and intention/uptake. Parents who identified as Democrat compared to Republican had greater confidence in the vaccines and had higher odds of vaccine intention/uptake for their children. Although rurality has been associated with vaccine confidence in the past we did not find that in our study. Instead, political affiliation appeared to explain most of the variation in vaccine confidence and intention/uptake, suggesting that more research is needed to identify best practices for using social media to reach parents with different political beliefs.

Governmental Persuasion Strategies on Social Media during COVID-19: A Comparative Study of the US and China

This study compared persuasive strategies of the governments of the U.S. and China during a public health crisis using social media messages. Collecting data with R and Python from two national public health sectors' official accounts on Twitter (N = 1,630) and Sina Weibo (N = 3,554), the researcher investigated how the organizations' messages reflected Cialdini's seven principles of persuasion and whether other emergent messaging patterns occurred. According to the different phases that the two countries had gone through during the pandemic, the researcher also conducted a pooled times series analysis to investigate the relationship between the frequency of daily posts and the number of daily COVID-19 positive new cases in the two countries. The study found that the principle of authority was the most often used rule by the two countries, and a combination of directive and non-directive messages was detected. The research discussed the effectiveness of Cialdini's principles in an online context and provided recommendations regarding timely responses towards the development of the disease on social media, which may also help build up the organizations' credibility in public health crises.

Mind the Gap: Exploring Nutritional Health Compared With Weight Management Interests of Individuals with Osteoarthritis

BackgroundFor persons with osteoarthritis (OA), nutrition education may facilitate weight and OA symptom management. ObjectivesThe primary aim of this study was to determine preferred OA-related nutritional and weight management topics and their preferred delivery modality. The secondary aim was to determine whether there is a disconnect between what patients want to know about nutrition and OA management and what information health-care professionals (HCPs) are providing to patients. MethodsThe Osteoarthritis Action Alliance surveyed individuals with OA to identify their preferences, categorized in 4 domains: 1) strategies for weight management and a healthy lifestyle; 2) vitamins, minerals, and other supplements; 3) foods or nutrients that may reduce inflammation; and 4) diets for weight loss. HCPs were provided these domains and asked which topics they discussed with patients with OA. Both groups were asked to select currently utilized or preferred formats of nutritional resources. ResultsSurvey responses from 338 individuals with OA and 104 HCPs were included. The highest preference rankings in each domain were: 1) foods that make OA symptoms worse (65%), foods and nutrients to reduce inflammation (57%), and healthy weight loss (42%); 2) glucosamine (53%), vitamin D (49%), and omega-3 fatty acids (45%); 3) spices and herbs (65%), fruits and vegetables (58%), and nuts (40%); and 4) Mediterranean diet (21%), low-carbohydrate diet (18%), and fasting or intermittent fasting (15%). There was greater than 20% discrepancy between interests reported by individuals with OA and discussions reported by HCPs on: weight loss strategies, general information on vitamins and minerals, special dietary considerations for other conditions, mindful eating, controlling caloric intake or portion sizes, and what foods worsen OA symptoms. Most respondents preferred to receive nutrition information in a passive format and did not want information from social media messaging. ConclusionsThere is disparity between the nutrition education content preferred by individuals with OA (which often lacks empirical support) and evidence-based topics being discussed by HCPs. HCPs must communicate evidence-based management of joint health and OA symptoms in patient-preferred formats. This study explored the information gap between what individuals with OA want to know and what HCPs believe they need to know.

Physique brand identity persuasiveness in social media message content

This study focuses on how Malaysian micro SMEs, in this case, food truck (FT) businesses, could digitally utilize social media as an effective branding tool for their products and services and create significant customer value to survive in the local competitive market. It is crucial for Malaysian SMEs (food truck, FT) operators to convince their customers through social media message content, specifically on what is being posted on their social media page. Thus, we examined the different aspects of persuasive message content taking place on social media. A qualitative approach using semi-structured interviews was employed to explore the SME food and beverage service industry. In addition, interviews were conducted with independent food truck founders and owners in Kuala Lumpur, Malaysia. This study yielded results that showed respondents' interest in physique brand identity persuasion in social media platforms. Creating an excellent physique brand identity persuasion recognized by its customers through social media has become an important activity for food truck (FT) owners to communicate their brand digitally.

Putting the blame back on Brussels: strategic communication of the populist radical right in the 2019 European Parliament elections

ABSTRACT The 2019 European Parliament elections were marked by a specific ‘threat to Europe’ discourse associated with concerns about possible populist radical right’s (PRR) breakthrough. The pro-European narrative suggested authoritarianism, non-democratic behaviour, anti-Europeanism, nostalgia for the past, and xenophobia as themes attributive of the PRR challenge. This article examines the PRR parties’ counter-narrative on these themes in the way it is embedded in their strategic communication during the election campaign. Three prominent examples are taken into consideration: France’s Rassemblement national, Italy’s Lega and Germany’s Alternative für Deutschland. By analysing discursively their election manifestos and social media messages the article outlines their claim for appropriation of the ‘threat to Europe’ discourse and European legitimacy of their Euroscepticism. Thus, the article seeks to contribute to the discussion on the ideological consolidation of the PRR.