Social Media Data Research Articles

The circular economy through the prism of machine learning and the YouTube video media platform

The transition to a Circular Economy (CE) is rapidly gaining ground across countries and industries. It is the means of achieving more sustainable development by adopting innovative environmentally friendly strategies and saving primary resources. There are several studies indicating the increasing public and corporate interest in the CE but still remain limited in terms of the multitude and utilization of social media data. This work aims to shed light on the most common topics discussed on the YouTube platform, related to the CE. For this reason, we selected 17 videos including the term “Circular Economy” since these have been the most relevant with a sufficient number of comments and views. The model identified two main topics referring to “Sustainable industry and environmental responsibility” and “Circular Economy and resource management” which is a strong indicator of the people's interest in the utilization of resources alongside industrial and corporate activities. The two-topic configuration presented the highest coherence score; however, five and ten-topic configurations have been deployed since there was no extreme differentiation in the model's performance, which could provide more detailed insights. This work's innovation lies in utilizing Machine Learning techniques and social media data to unravel CE's debates.

Too much focus on your health might be bad for your health: Reddit user's communication style predicts their Long COVID likelihood.

Long Covid is a chronic disease that affects more than 65 million people worldwide, characterized by a wide range of persistent symptoms following a Covid-19 infection. Previous studies have investigated potential risk factors contributing to elevated vulnerability to Long Covid. However, research on the social traits associated with affected patients is scarce. This study introduces an innovative methodological approach that allows us to extract valuable insights directly from patients' voices. By analyzing written texts shared on social media platforms, we aim to collect information on the psychological aspects of people who report experiencing Long Covid. In particular, we collect texts of patients they wrote BEFORE they were afflicted with Long Covid. We examined the differences in communication style, sentiment, language complexity, and psychological factors of natural language use among the profiles of 6.107 Reddit users, distinguishing between those who claim they have never contracted Covid -19, those who claim to have had it, and those who claim to have experienced Long Covid symptoms. Our findings reveal that people in the Long Covid group frequently discussed health-related topics before the pandemic, indicating a greater focus on health-related concerns. Furthermore, they exhibited a more limited network of connections, lower linguistic complexity, and a greater propensity to employ emotionally charged expressions than the other groups. Using social media data, we can provide a unique opportunity to explore potential risk factors associated with Long Covid, starting from the patient's perspective.

Fine-grained sentiment analysis using multidimensional feature fusion and GCN

ABSTRACT The demand for analysing sentiment information in social media data is increasing. However, current fine-grained sentiment analysis methods fail to consider both global and local semantic features simultaneously, leading to the oversight of grammatical information in sentences and an inability to address the issue of polysemy. To address these challenges, we propose a microblog fine-grained sentiment analysis model based on multidimensional feature fusion and graph convolutional neural networks (GCN). Built upon the ALBERT model, we utilize BiLSTM and capsule network models to extract global and local semantic features, thereby capturing bidirectional semantic dependencies and textual positional semantic information. Finally, we employ multi-head self-attention and GCN to select key features and sentence information, ensuring the integrity of fine-grained features. The experimental results indicate that the model outperforms several other models on fine-grained sentiment analysis datasets SMP2020-EWECT, NLPCC2013, NLPCC2014, and the binary classification dataset weibo_senti_100k, achieving accuracies of 80.64%, 67.19%, 71.37%, and 98.43%, respectively.

An analysis of public topics and sentiments based on social media during the COVID-19 Omicron Variant outbreak in Shanghai 2022

The outbreak of the COVID-19 Omicron variant in Shanghai in 2022 elicited complex emotions among Shanghainese during the two-month quarantine period. This paper aims to identify prevailing public themes and sentiments by analyzing social media posts from Weibo. Initially, we conducted research based on a dataset of 90,000 Weibo posts during the 2022 COVID-19 outbreak in Shanghai. By examining social media data that mirrors residents' emotional shifts and areas of focus during unforeseen circumstances, we have developed an analytical framework combining hotspot analysis and public sentiment assessment. Subsequently, we employed the Latent Dirichlet Allocation (LDA) method to conduct topic modeling on the Weibo text data. The SnowNLP sentiment classification method was then utilized to quantify sentiment values. Ultimately, we performed spatial visualization of sentiment and concern data, categorizing them into distinct time periods based on Shanghai's infection curve. This approach allowed us to investigate concern focal points, sentiment trends, and their spatiotemporal evolution characteristics. Our findings indicate that variations in public sentiment primarily hinge on the severity of the epidemic's spread, emerging events, the availability of essential resources, and the government's ability to respond promptly and accurately. It is evident that, while residents' concerns shift over time, their primary objective on social media remains expressing demands and releasing emotions. This research offers an avenue for leveraging public opinion analysis to enhance governance capacity during crises, fortify urban resilience, and promote public involvement in governmental decision-making processes.

Extracting geo-references from social media text using bi-long short term memory networks

<p>The social media data provides great source of information about global and local events, with millions of users. More precisely, the fact that brief messages are practical and are highly popular. Many recent studies have been motivated to estimate the location of the events identified by tracking posts in social media text messages. It might be difficult to extract location data and estimate the location of an event while maintaining a sufficient level of situation awareness, particularly in disaster situations like fires or traffic accidents. In this presented work we proposed an approach to identify geo-references in the text messages. We used bi-directional long short term memory (LSTM) neural networks to extract location information in the text messages. The results show that applying Bi-LSTM on dataset gives high level accuracy after fine-tuning (up to 10 epochs). The testing results show that accuracy achieved is 0.98 and 0.076 loss value. This proves that the proposed methodology is better than the previous conditional random field (CRF)-based approaches.</p>

The Influence of Social Media Emotions: Evidence from the Kodak Manic Episode

This paper explores the impact of Kodak’s announcements of ventures into the cryptocurrency and pharmaceutical industries on market sentiment and investor emotions. Evaluating social media data using novel sentiment analysis techniques, we find that emotions such as excitement, mania, anxiety, panic, and revulsion were salient in response to the announcements of KodakCoin and KodakPharma. Our results suggest that these emotions had significant short-term effects on share prices and market perception, demonstrating the impact of emotional patterns observed within KodakCoin and KodakPharma events on investor behavior and emphasizing the vulnerability of unsophisticated online stock traders to corporate ‘pivots’ irrespective of the perceived associated risks. We highlight the significance of monitoring emerging technologies and balancing innovation with stability, contributing to developing a more resilient financial ecosystem. Furthermore, this research underscores the importance of addressing several key areas in policy and regulation to ensure a stable and transparent financial ecosystem.

Using Social Media Data to Understand Citizen Perceptions of Urban Planning in a City Simulation Game

Background City simulation games provide players a gaming experience by simulating different aspects of the real city. While there is an increasing scholarly interest in games for social learning and education, little research has been conducted to understand citizen perceptions and understanding of urban planning issues in city simulation games. Aim This study aims to understand the affective perception and cognitive learning of citizens regarding urban planning elements in the online communities of Cities: Skylines. Research Methods We develop a new methodological approach based on social media data analytics. Large datasets were scraped from Reddit, the most popular social media platform for video game players. The collected data were subjected to content analysis and sentiment analysis that identify different types of topics and emotions to understand citizens’ cognitive and affective perspectives. Key Findings and Conclusion The findings show that positive emotions were often about the game design, while negative emotions conveyed real-world planning problems such as transportation concerns. The cognitive dimension uncovered citizens’ urban recognition tied to personal experiences in various geographical contexts. This study has practical implications for game design for urban planning.

Natural Language Processing (NLP) for Sentiment Analysis in Social Media

Purpose: This study sought to analyze Natural Language Processing (NLP) for sentiment analysis in social media. Methodology: The study adopted a desktop research methodology. Desk research refers to secondary data or that which can be collected without fieldwork. Desk research is basically involved in collecting data from existing resources hence it is often considered a low cost technique as compared to field research, as the main cost is involved in executive’s time, telephone charges and directories. Thus, the study relied on already published studies, reports and statistics. This secondary data was easily accessed through the online journals and library. Findings: The findings reveal that there exists a contextual and methodological gap relating to Natural Language Processing (NLP) for sentiment analysis in social media. Preliminary empirical review revealed that advanced computational techniques significantly advanced the understanding of sentiments across diverse social media platforms. Methodologies such as machine learning algorithms and deep learning models like CNNs and RNNs demonstrated robust capabilities in categorizing sentiments accurately and capturing contextual nuances such as sarcasm and slang. The research highlighted the interdisciplinary nature of NLP applications, integrating linguistics with computer science and social sciences to develop effective frameworks for analyzing large-scale social media data. These findings contributed to enhancing decision-making in marketing, politics, and public opinion research, pointing towards future directions in hybrid NLP models for improved sentiment analysis across different languages and cultural contexts. Unique Contribution to Theory, Practice and Policy: The Social Constructionism, Cognitive Linguistics and Discourse Analysis Theory may be used to anchor future studies on Natural Language Processing (NLP). The recommendations aimed to advance theoretical foundations by exploring deep learning models and nuanced sentiment lexicons. Practical applications were enhanced through the development of scalable NLP tools for real-time data processing and integration into social media platforms. Policy implications focused on establishing ethical guidelines for data privacy and bias mitigation in sentiment analysis algorithms. Cross-disciplinary collaboration fostered innovation by integrating insights from computer science, linguistics, psychology, and social sciences. Education initiatives and international collaborations were prioritized to build capacity and standardize methodologies globally, ensuring advancements in both research and practical deployment of sentiment analysis technologies.

Data Privacy in Social Media: Awareness and Comfortability of Mathematical Computing Course Session I 2023/2024 Sultan Abdul Halim Mu'adzam Shah Polytechnic’s Students

The increasing integration of social media into daily lives of individuals has raised significant concerns about data privacy. As users actively engage with various platforms, the extent to which they are aware of the privacy implications and their comfort levels with the use of personal data become vital issues. This study aimed to identify social media platforms that are often used by students, their level of awareness towards social media data privacy policy and their comfortability towards data privacy by social media companies. The respondents in this study comprises of 106 male and female students. The data are analyzed descriptively by mean, frequency and percentage. The level of awareness towards social media data privacy policy is high with average mean score 3.94, WhatsApp is the often-surfed social media platform by score of 89.62% and most students were highly comfortable with how educational social media platform used their data with mean 3.85. This study proposes the necessity of a diverse approach involving education, user empowerment, and technological advancements to enhance data privacy awareness and comfort levels among students in the realm of social media and prepare to a more secure and responsible online environment for the next generation users

Tweeting the academic resistance in Turkey and Hungary: from cultural crisis to defending the nation

ABSTRACT This paper explores the reemergence of universities as spaces for the contestation of government policies in an era of illiberalism, rising nationalism and populism. We focus on high-profile contestations occurring at two public universities in Turkey and Hungary: Boğaziçi University and the University of Theatre and Film Arts, respectively. At both universities, attacks on democratic governance and liberal-democratic values have resulted in academic resistance that have played out on physical campuses and in online spaces. We conceptualise a virtual space as a space of contestation and analyse social media data to identify the dominant narratives of governments and their supporters on one side and resistance supporters on the other that gives the way to two major narratives: while right-wing imaginaries and pro-government actors frame the liberal and pro-freedom inclinations of universities as a ‘threat’ to the nation, students and faculty portray the government as the true threat to their rights and freedoms and knowledge production within what was once a ‘safe space’ in universities, thus framing themselves as the ‘true guardians’ of their institutions. Our analysis shows how universities have become contemporary flashpoints in and upon which broader political contestations over national culture play out.

A comprehensive analysis of LSTM techniques for predicting financial market

The “Financial Engineering” has revolutionized the financial industry by integrating mathematics, finance, economics, statistics, and computational tools to solve complex problems like risk management and portfolio optimization. This interdisciplinary approach has given rise to “AI in Finance”, merging the quantitative techniques of financial engineering with AI’s data-driven capabilities. In this survey, we explore concrete examples of their real-world impact. One such example involves a leading investment firm using machine learning algorithms to analyze market sentiments for informed trading decisions, resulting in significant returns. Additionally, AI-driven credit scoring models are expanding financial access by accurately assessing creditworthiness, especially for underserved populations. Natural language processing algorithms are also employed to parse financial news and social media data, providing investors with timely insights to navigate volatile markets effectively. These advancements highlight the transformative potential of financial engineering and AI in finance. By optimizing investment strategies and mitigating risks, they drive innovation and resilience in today’s dynamic financial landscape. From algorithmic trading to credit risk assessment and market sentiment analysis, the fusion of these disciplines is reshaping traditional paradigms and shaping the future of finance.

Customers’ metaverse service encounter perceptions: sentiment analysis and topic modeling

ABSTRACT Using machine learning, we examined customers’ opinions about the metaverse in the hospitality industry (encompassing hotels, restaurant, gaming, virtual events, tours and travel). A total of 8,855 tweets were collected from Twitter (now called X), and machine learning algorithms such as sentiment analysis and topic modeling were performed using Python libraries to capture the important topics related to metaverse applications. Nearly two thirds of the collected tweets (60.9%) contained a mostly positive general sentiment toward the use of the metaverse. Six important topics emerged from the topic modeling: gaming, virtual events, virtual sightseeing, travel, business and blockchain. Despite numerous studies on the proper integration of the metaverse, VR and AR, to the best of our knowledge, this is one of the first studies conducted to determine the customer experience of the metaverse in the hospitality industry using social media data.

Social media mining for business inequality: modified spotted hyena optimization based random forest

Businesses may advertise, communicate with customers and research industry trends on social media today. Due to the massive amount of data generated on social media platforms, organizations struggle to acquire valuable data and fix marketing inequalities. To combine Modified Spotted Hyena Optimization with Random Forest (MSHO-RF) for classification in social media mining for a business inequality study. Data-driven solutions to inequality in operations and customer interactions should increase prediction accuracy, lower feature dimensionality and extract actionable insights. We begin by collecting social media mining for business inequality data. Afterward, Adaptive mean filter (AMF) techniques were used to initialize the data preprocessing. With the use of Linear Discriminant Analysis (LDA), we examined the most suitable arrangement of feature extraction. This research offers a unique method MSHO-RF, to deal with the problem of business inequality using social media mining. The MSHO-RF algorithm had higher results than compared to a variety of performance parameters, featuring a 94% of F1-score, a precision rate of 97%, an accuracy rate of 98% and recall rates of 95%. Our findings suggest that firms might benefit from a framework for leveraging social media data to promote more equitable and sustainable practices if they combine MSHO-RF with social media mining for business inequality. This approach benefits companies and their stakeholders by fostering an open, data-driven ecosystem.

Passive citizen science: social media as a tool for marine wildlife observation

It is not new that species occurrence records provided by citizens are valuable contributions to marine research. However, passive citizen science is a recent and promising methodology of crowdsourcing. Here, we review existing studies that used spontaneous posts shared by users on social media to obtain marine wildlife scientific data. We also present a critical analysis of passive citizen science using social media. Specifically, we (1) introduce marine citizen science and (2) analyze the advantages of passive citizen science compared with voluntary approaches. We also (3) present investigations that extracted information and produced knowledge for marine ecology and conservation through passive citizen science, (4) discuss best practices and opportunities, and (5) identify challenges of this approach. Although social media data may have inherent biases and diverse ethical issues, the data volume is generally large, and the benefits of obtaining observations at a low cost and in real time compensate for some shortcomings, which can sometimes be mitigated. Considering that the use of passive citizen science will continue to grow and be part of our daily lives, we expect the current review to be useful for future investigations.

An integrated framework for flood disaster information extraction and analysis leveraging social media data: A case study of the Shouguang flood in China

Flood disasters cause significant casualties and economic losses annually worldwide. During disasters, accurate and timely information is crucial for disaster management. However, remote sensing cannot balance temporal and spatial resolution, and the coverage of specialized equipment is limited, making continuous monitoring challenging. Real-time disaster-related information shared by social media users offers new possibilities for monitoring. We propose a framework for extracting and analyzing flood information from social media, validated through the 2018 Shouguang flood in China. This framework innovatively combines deep learning techniques and regular expression matching techniques to automatically extract key flood-related information from Weibo textual data, such as problems, floodings, needs, rescues, and measures, achieving an accuracy of 83 %, surpassing traditional models like the Biterm Topic Model (BTM). In the spatiotemporal analysis of the disaster, our research identifies critical time points during the disaster through quantitative analysis of the information and explores the spatial distribution of calls for help using Kernel Density Estimation (KDE), followed by identifying the core affected areas using the Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) algorithm. For semantic analysis, we adopt the Latent Dirichlet Allocation (LDA) algorithm to perform topic modeling on Weibo texts from different regions, identifying the types of disasters affecting each township. Additionally, through correlation analysis, we investigate the relationship between disaster rescue requests and response measures to evaluate the adequacy of flood response measures in each township. The research results demonstrate that this analytical framework can accurately extract disaster information, precisely identify critical time points in flood disasters, locate core affected areas, uncover primary regional issues, and further validate the sufficiency of response measures, therefore enhancing the efficiency in collecting disaster information and analytical capabilities.

Changes in online public opinions associated with various three-child supportive policies in China: Observational study using social media data over time.

According to the Seventh National Census, China's fertility rate is less than 1.5, marking a significant national issue with potential risks. To counter this low birth rate, the Chinese government has relaxed family planning policies and introduced supportive measures. Changes in birth policy have attracted considerable attention from the people of China. This article aims to study the public's response to the three-child support policy using Weibo as a window. The goal is to provide a more balanced evaluation of current perspectives, enabling policymakers to formulate better fertility information, particularly when anticipating a poor public response to controversial policies. This research uses a crawler to gather data from Sina Weibo. Through opinion mining of Weibo posts on the three-child policy, Weibo users' online opinions on the three-child policy are analyzed from two perspectives: their attention content and sentiment tendency. Using an interrupted time series, it examines changes in online views on the policy, matching policy documents to the time nodes of Weibo posts. The public has shown great interest in and provided short-term positive feedback on policies related to improving maternity insurance, birth rewards, and housing subsidies. In contrast, there has been a continuous negative response to policies such as extending maternity leave, which has particularly sparked concerns among women regarding future employment and marital rights protection. On social media, the public's attention to the three-child birth policy has focused mainly on the protection of women's rights, especially legal rights after childbirth, and issues related to physical and mental health. Child-rearing support and economic pressure are also hot topics, involving the daily expenses of multichild families, childcare services, and housing pressure. However, this study also revealed that infertile or single women express a strong desire to have children, but due to limitations in the personal medical insurance system, this desire has not been fully satisfied. Our study demonstrates the feasibility of a rapid and flexible method for evaluating the public response to various three-child supportive policies in China using near real-time social media data. This information can help policy makers anticipate public responses to future pandemic three-child policies and ensure that adequate resources are dedicated to addressing increases in negative sentiment and levels of disagreement in the face of scientifically informed but controversial, restrictions.

Carbon emissions forecasting based on tensor decomposition with multi-source data fusion

Accurately forecasting carbon dioxide emissions is crucial for policymakers and researchers aiming to combat climate change and develop effective emission reduction strategies. This study introduces an innovative method that leverages multi-source social media information to address the challenges of insufficient information and data uncertainty in carbon emission time series forecasting. We propose a combined Tensor-LSTM-ARIMA model for predicting carbon emissions, utilizing tensor decomposition data analysis methods. The results indicate that this combined model effectively captures the complex relationships within heterogeneous data, outperforming baseline models in prediction accuracy. Furthermore, the study demonstrates that unstructured social media data can enhance structured time series data, providing a new perspective for comprehensively understanding the variables influencing carbon emission predictions.

Data Analytics for Predicting Situational Developments in Smart Cities: Assessing User Perceptions.

The analysis of large volumes of data collected from heterogeneous sources is increasingly important for the development of megacities, the advancement of smart city technologies, and ensuring a high quality of life for citizens. This study aimed to develop algorithms for analyzing and interpreting social media data to assess citizens' opinions in real time and for verifying and examining data to analyze social tension and predict the development of situations during the implementation of urban projects. The developed algorithms were tested using an urban project in the field of transportation system development. The study's material included data from social networks, messenger channels and chats, video hosting platforms, blogs, microblogs, forums, and review sites. An interdisciplinary approach was utilized to analyze the data, employing tools such as Brand Analytics, TextAnalyst 2.32, GPT-3.5, GPT-4, GPT-4o, and Tableau. The results of the data analysis showed identical outcomes, indicating a neutral perception among users and the absence of social tension surrounding the project's implementation, allowing for the prediction of a calm development of the situation. Additionally, recommendations were developed to avert potential conflicts and eliminate sources of social tension for decision-making purposes.

Perception study of urban green spaces in Singapore urban parks: Spatio-temporal evaluation and the relationship with land cover

In the current era of increasing urbanization, urban green spaces play a crucial role in enhancing human well-being. However, quantifying public perceptions from text data at spatio-temporal scales remains challenging, and the relationship between urban green space perception and spatial-physical attributes requires further exploration. This study systematically examines public perceptions of urban green spaces within Singapore's urban parks from 2018 to 2022. Utilizing Twitter data, it applies large language models to conduct textual content analysis related to urban green space. The findings reveal a positive trend, with individuals expressing favorable perceptions and satisfaction towards urban green spaces in Singapore. Specifically, this study demonstrates that people's perceptions of urban green spaces are influenced by vegetation density. Higher vegetation density heightens people's awareness of spatial presence, while shrub and grassland may lead to neglect of urban green spaces as individuals focus more on themselves. Additionally, due to the spatial heterogeneity of the area, there is no clear correlation between all land covers and public satisfaction with urban green spaces in Singapore. The results also indicate a significant decrease in public perception in 2020, followed by a subsequent recovery. This fluctuation is attributed to the substantial impact of the COVID-19 pandemic, suggesting that external socio-political, economic, and public health events can impact public green space needs and spatial perceptions. In conclusion, this study contributes to the understanding of urban green spaces by effectively analyzing textual content extracted from social media data using large language models. The insights gained contribute valuable to the following discussions regarding the planning and design of urban green spaces and urban parks.

Research Paper on Artificial Intelligence

Artificial Intelligence (AI) has emerged as a transformative force across various sectors, revolutionizing processes, enhancing efficiency, and redefining innovation. This research paper delves into the multifaceted landscape of AI, focusing on its applications, knowledge representation, and implications for innovation. The paper begins by exploring the diverse applications of AI across healthcare, gaming, finance, data security, social media, robotics, and e-commerce. In healthcare, AI aids in diagnosis and patient care, while in gaming, it enables strategic game play and enhances user experience. The finance sector leverages AI for automation, analytics, and algorithmic trading, improving decision-making and customer service. AI also plays a vital role in ensuring data security through advanced detection systems, manages vast social media data for enhanced user engagement, and drives innovation in robotics and e-commerce. Moving forward, the paper delves into the realm of expert systems and knowledge representation, elucidating the role of AI in simulating human expertise and modeling complex information structures. It discusses various aspects of knowledge representation, such as propositional knowledge representation, image retrieval, functional relationships between objects, and class representation formalism, highlighting their significance in developing intelligent systems. Furthermore, the paper examines the integration of AI in maintenance practices, both for tangible systems like engineering workshops and intangible products like data extraction wrappers. It underscores the importance of AI in optimizing operational efficiency, reducing downtime, and ensuring continuous data extraction. Lastly, the paper explores the concept of deep learning as a general- purpose invention, discussing its potential implications for innovation, management, institutions, and policy. It addresses key issues such as the management and organization of innovation, intellectual property rights, competition policy, and the cumulative knowledge production facilitated by deep learning. In conclusion, this research paper provides a comprehensive overview of AI's transformative potential, emphasizing the need for further research and analysis to fully comprehend its impact on society, economy, and innovation.