Search Results (181)

In recent years, the proliferation of fake news and misinformation has grown exponentially, far surpassing that of genuine news and posing a serious threat to social stability. Existing research in fake news detection primarily applies contrastive learning methods with a single-hot labeling strategy. The issue does not lie with contrastive learning as a technique but with its current application in fake news detection systems. Specifically, these systems penalize all negative samples equally due to the use of single-hot labeling, thus overlooking the underlying semantic relationships among negative samples. As a result, contrastive learning models tend to learn from simple samples while neglecting highly deceptive samples located at the boundary between true and false, as well as the heterogeneity of text-image features, which complicates cross-modal fusion. To mitigate these known limitations in current applications, this paper proposes a fake news detection method based on contrastive learning and cross-modal interaction. First, a consistency-aware soft-label contrastive learning mechanism based on semantic similarity is designed to provide more granular supervision signals for contrastive learning. Secondly, a difficult negative sample mining strategy based on a similarity matrix is designed to optimize the symmetry alignment of image and text features, which effectively improves the model’s ability to discriminate boundary samples. To further optimize the feature fusion process, a cross-modal interaction module is designed to learn the symmetric interaction relationship between image and text features. Finally, an attention mechanism is designed to adaptively adjust the contributions of text-image features and interaction features, forming the final multimodal feature representation. Experiments are conducted on two major social media platform datasets, and compared with existing methods, the proposed method effectively improves the detection capability of fake news. Full article

This study focuses on cultural and urban development policies under China’s 14th Five-Year Plan, exploring the content and semantic structure of discussions on the “East Asian Cultural Capital” project on the Weibo platform. It analyzes how national cultural development policies are reflected in the discourse system related to the “East Asian Cultural Capital” on social media and emphasizes the guiding role of policies in the dissemination of online culture. When China announced the 14th Five-Year Plan in 2021, the strategic direction and policy framework for cultural development over the five-year period from 2021 to 2025 were clearly outlined. This study employs text mining and semantic network analysis methods to analyze user-generated content on Weibo from 2023 to 2024, aiming to understand public perception and discourse trends. Word frequency and TF-IDF analyses identify key terms and issues, while centrality and CONCOR clustering analyses reveal the semantic structure and discourse communities. MR-QAP regression is employed to compare network changes across the two years. Findings highlight that urban cultural development, heritage preservation, and regional exchange are central themes, with digital media, cultural branding, trilateral cooperation, and cultural–economic integration emerging as key factors in regional collaboration. Full article

This study investigates the mechanisms underlying the diffusion of risk information about genetically modified organisms (GMOs) on the Chinese social media platform Weibo. Drawing upon social contagion theory, we examine how endogenous and exogenous mechanisms shape users’ information-sharing behaviors. An analysis of 388,722 reposts from 2444 original GMO risk-related texts enabled the construction of a comprehensive sharing network, with computational text-mining techniques employed to detect users’ attitudes toward GMOs. To bridge the gap between descriptive and inferential network analysis, we employ a Shannon entropy-based approach to quantify the uncertainty and concentration of attitudinal differences and similarities among sharing and non-sharing dyads, providing an information-theoretic foundation for understanding positional and differential homophily. The entropy-based analysis reveals that information-sharing ties are characterized by lower entropy in attitude differences, indicating greater attitudinal alignment among sharing users, especially among GMO opponents. Building on these findings, the Exponential Random Graph Model (ERGM) further demonstrates that both endogenous network mechanisms (reciprocity, preferential attachment, and triadic closure) and positional homophily influence GMO risk information sharing and dissemination. A key finding is the presence of a differential homophily effect, where GMO opponents exhibit stronger homophilic tendencies than non-opponents. Despite the prevalence of homophily, this paper uncovers substantial cross-attitude interactions, challenging simplistic notions of echo chambers in GMO risk communication. By integrating entropy and ERGM analyses, this study advances a more nuanced, information-theoretic understanding of how digital platforms mediate public perceptions and debates surrounding controversial socio-scientific issues, offering valuable implications for developing effective risk communication strategies in increasingly polarized online spaces. Full article

The rapid development and deployment of artificial intelligence (AI) technologies have sparked intense public interest and debate. While these innovations promise to revolutionise various aspects of human life, it is crucial to understand the complex emotional responses they elicit from potential adopters and users. Such findings can offer crucial guidance for stakeholders involved in the development, implementation, and governance of AI technologies like OpenAI’s ChatGPT, a large language model (LLM) that garnered significant attention upon its release, enabling more informed decision-making regarding potential challenges and opportunities. While previous studies have employed data-driven approaches towards investigating public reactions to emerging technologies, they often relied on sentiment polarity analysis, which categorises responses as positive or negative. However, this binary approach fails to capture the nuanced emotional landscape surrounding technological adoption. This paper overcomes this limitation by presenting a comprehensive analysis for investigating the emotional landscape surrounding technology adoption by using the launch of ChatGPT as a case study. In particular, a large corpus of social media texts containing references to ChatGPT was compiled. Text mining techniques were applied to extract emotions, capturing a more nuanced and multifaceted representation of public reactions. This approach allows the identification of specific emotions such as excitement, fear, surprise, and frustration, providing deeper insights into user acceptance, integration, and potential adoption of the technology. By analysing this emotional landscape, we aim to provide a more comprehensive understanding of the factors influencing ChatGPT’s reception and potential long-term impact. Furthermore, we employ topic modelling to identify and extract the common themes discussed across the dataset. This additional layer of analysis allows us to understand the specific aspects of ChatGPT driving different emotional responses. By linking emotions to particular topics, we gain a more contextual understanding of public reaction, which can inform decision-making processes in the development, deployment, and regulation of AI technologies. Full article

When a disaster occurs, a large number of social media posts on platforms like Weibo attract public attention with their combination of text and images. However, the consistency between textual descriptions and visual representations varies significantly. Consistent multi-modal data are crucial for helping the public understand the disaster situation and support rescue efforts. This study aims to develop a systematic framework for assessing the consistency of multi-modal disaster-related data on social media. This study explored how the congruence between text and image content affects public engagement and informs strategies for efficient emergency responses. Firstly, the Clip (Contrastive Language-Image Pre-Training) model was used to mine the disaster correlation, loss category, and severity of the images and text. Then, the consistency of image–text pairs was qualitatively analyzed and quantitatively calculated. Finally, the influence of graphic consistency on social concern was discussed. The experimental findings reveal that the consistency of text and image data significantly influences the degree of public concern. When the consistency increases by 1%, the social attention index will increase by about 0.8%. This shows that consistency is a key factor for attracting public attention and promoting the dissemination of information related to important disasters. The proposed framework offers a robust, systematic approach to analyzing disaster loss information consistency. It allows for the efficient extraction of high-consistency data from vast social media data sets, providing governments and emergency response agencies with timely, accurate insights into disaster situations. Full article

Understanding customer attitudes has become a critical component of decision-making due to the growing influence of social media and e-commerce. Text-based opinions are the most structured, hence playing an important role in sentiment analysis. Most of the existing methods, which include lexicon-based approaches and traditional machine learning techniques, are insufficient for handling contextual nuances and scalability. While the latter has limitations in model performance and generalization, deep learning (DL) has achieved improvement, especially on semantic relationship capturing with recurrent neural networks (RNNs) and convolutional neural networks (CNNs). The aim of the study is to enhance opinion mining by introducing a hybrid deep neural network model that combines a bidirectional gated recurrent unit (BGRU) and long short-term memory (LSTM) layers to improve sentiment analysis, particularly addressing challenges such as contextual nuance, scalability, and class imbalance. To substantiate the efficacy of the proposed model, we conducted comprehensive experiments utilizing benchmark datasets, encompassing IMDB movie critiques and Amazon product evaluations. The introduced hybrid BGRU-LSTM (HBGRU-LSTM) architecture attained a testing accuracy of 95%, exceeding the performance of traditional DL frameworks such as LSTM (93.06%), CNN+LSTM (93.31%), and GRU+LSTM (92.20%). Moreover, our model exhibited a noteworthy enhancement in recall for negative sentiments, escalating from 86% (unbalanced dataset) to 96% (balanced dataset), thereby ensuring a more equitable and just sentiment classification. Furthermore, the model diminished misclassification loss from 20.24% for unbalanced to 13.3% for balanced dataset, signifying enhanced generalization and resilience. Full article

In this study, we propose a computational approach that applies text mining and deep learning to conduct controversy detection on social media platforms. Unlike previous research, our method integrates multidimensional and heterogeneous information from social media into a heterogeneous signed attributed network, encompassing various users’ attributes, semantic information, and structural heterogeneity. We introduce a deep dual-layer self-supervised algorithm for community detection and analyze controversy within this network. A novel controversy metric is devised by considering three dimensions of controversy: community distinctions, betweenness centrality, and user representations. A comparison between our method and other classical controversy measures such as Random Walk, Biased Random Walk (BRW), BCC, EC, GMCK, MBLB, and community-based methods reveals that our model consistently produces more stable and accurate controversy scores. Additionally, we calculated the level of controversy and computed p-values for the detected communities on our crawled dataset Weibo, including #Microblog (3792), #Comment (45,741), #Retweet (36,126), and #User (61,327). Overall, our model had a comprehensive and nuanced understanding of controversy on social media platforms. To facilitate its use, we have developed a user-friendly web server. Full article

The rapid development of social media has driven the need for opinion mining and sentiment analysis based on multimodal samples. As a fine-grained task within multimodal sentiment analysis, aspect-based multimodal sentiment analysis (ABMSA) enables the accurate and efficient determination of sentiment polarity for aspect-level targets. However, traditional ABMSA methods often perform suboptimally on social media samples, as the images in these samples typically contain embedded text that conventional models overlook. Such text influences sentiment judgment. To address this issue, we propose a text-in-image enhanced self-supervised alignment model (TESAM) that accounts for multimodal information more comprehensively. Specifically, we employed Optical Character Recognition technology to extract embedded text from images and, based on the principle that text-in-image is an integral part of the visual modality, fused it with visual features to obtain more comprehensive image representations. Additionally, we incorporate aspect words to guide the model in disregarding irrelevant semantic features, thereby reducing noise interference. Furthermore, to mitigate the semantic gap between modalities, we propose pre-training the feature extraction module with self-supervised alignment. During this pre-training stage, unimodal semantic embeddings from both modalities are aligned by calculating errors using Euclidean distance and cosine similarity. Experimental results demonstrate that TESAM achieved remarkable performances on three ABMSA benchmarks. These results validate the rationale and effectiveness of our proposed improvements. Full article

Urban green spaces (UGSs) provide recreational and cultural services to urban residents and play an important role in mental health. This study uses big data mining techniques to analyze 62 urban parks in the Changsha–Zhuzhou–Xiangtan urban agglomeration (CZXUA) based on data such as points of interest (POIs), areas of interest (AOIs), and user comments from the popular social media platform Dianping. In addition, the authors apply sentiment analysis using perceptual dictionaries combined with geographic information data to identify text emotions. A structural equation model (SEM) was constructed in IBM SPSS AMOS 24.0 software to investigate the relationship between five external features, five types of cultural services, nine landscape elements, four environmental factors, and tourist emotions. The results show that UGS external features, cultural services, landscape elements, and environmental factors all have positive effects on residents’ emotions, with landscape elements having the greatest impact. The other factors show similar effects on residents’ moods. In various UGSs, natural elements such as vegetation and water tend to evoke positive emotions in residents, while artificial elements such as roads, squares, and buildings elicit more varied emotional responses. This research provides science-based support for the design and management of urban parks. Full article

The display methods for traditional cultural heritage lack personalization and emotional interaction, making it difficult to stimulate the public’s deep cultural awareness. This is especially true in commercialized historical districts, where cultural value is easily overlooked. Balancing cultural value and commercial value in information display has become one of the challenges that needs to be addressed. To solve the above problems, this article focuses on the identification of deep cultural values and the optimization of the information display in Beijing’s Qianmen Street, proposing a framework for cultural information mining and display based on affective computing and large language models. The pre-trained models QwenLM and RoBERTa were employed to analyze text and image data from user-generated content on social media, identifying users’ emotional tendencies toward various cultural value dimensions and quantifying their multilayered understanding of architectural heritage. This study further constructed a multimodal information presentation model driven by emotional feedback, mapping it into virtual reality environments to enable personalized, multilayered cultural information visualization. The framework’s effectiveness was validated through an eye-tracking experiment that assessed how different presentation styles impacted users’ emotional engagement and cognitive outcomes. The results show that the affective computing and multimodal data fusion approach to cultural heritage presentation accurately captures users’ emotions, enhancing their interest and emotional involvement. Personalized presentations of information significantly improve users’ engagement, historical understanding, and cultural experience, thereby fostering a deeper comprehension of historical contexts and architectural details. Full article

Sarcasm detection is a crucial task in natural language processing (NLP), particularly in sentiment analysis and opinion mining, where sarcasm can distort sentiment interpretation. Accurately identifying sarcasm remains challenging due to its context-dependent nature and linguistic complexity across informal text sources like social media and conversational dialogues. This study utilizes three benchmark datasets, namely, News Headlines, Mustard, and Reddit (SARC), which contain diverse sarcastic expressions from headlines, scripted dialogues, and online conversations. The proposed methodology leverages transformer-based models (RoBERTa and DistilBERT), integrating context summarization, metadata extraction, and conversational structure preservation to enhance sarcasm detection. The novelty of this research lies in combining contextual summarization with metadata-enhanced embeddings to improve model interpretability and efficiency. Performance evaluation is based on accuracy, F1 score, and the Jaccard coefficient, ensuring a comprehensive assessment. Experimental results demonstrate that RoBERTa achieves 98.5% accuracy with metadata, while DistilBERT offers a 1.74x speedup, highlighting the trade-off between accuracy and computational efficiency for real-world sarcasm detection applications. Full article

In the digital age, the rapid proliferation of misinformation and disinformation poses a critical challenge to societal trust and the integrity of public discourse. This study presents a comprehensive machine learning framework for fake news detection, integrating advanced natural language processing techniques and deep learning architectures. We rigorously evaluate a diverse set of detection models across multiple content types, including social media posts, news articles, and user-generated comments. Our approach systematically compares traditional machine learning classifiers (Naïve Bayes, SVMs, Random Forest) with state-of-the-art deep learning models, such as CNNs, LSTMs, and BERT, while incorporating optimized vectorization techniques, including TF-IDF, Word2Vec, and contextual embeddings. Through extensive experimentation across multiple datasets, our results demonstrate that BERT-based models consistently achieve superior performance, significantly improving detection accuracy in complex misinformation scenarios. Furthermore, we extend the evaluation beyond conventional accuracy metrics by incorporating the Matthews Correlation Coefficient (MCC) and Receiver Operating Characteristic–Area Under the Curve (ROC–AUC), ensuring a robust and interpretable assessment of model efficacy. Beyond technical advancements, we explore the ethical implications of automated misinformation detection, addressing concerns related to censorship, algorithmic bias, and the trade-off between content moderation and freedom of expression. This research not only advances the methodological landscape of fake news detection but also contributes to the broader discourse on safeguarding democratic values, media integrity, and responsible AI deployment in digital environments. Full article

Digital transformation (DT) has become an important engine for the development of the digital economy and an important means of reshaping corporate culture, business processes, management models, and so on. Different social communities at different levels have different needs and understandings of digital transformation. Therefore, this paper proposes to explore the communication themes of digital transformation on social media. This study’s main objective is to uncover underlying thematic structures and core ideas from large amounts of textual data in different social media communities to better understand the significance of the communication themes. This paper also aims to reveal the characteristics of diffusion patterns of DT themes by opinion-themed mining. This study uses text mining and social network analysis methods to mine DT themes, theme structure, and the statistical characteristics of hot words across various online communities. The main findings of this study are as follows. The Huawei forum discusses the technological drivers of the digital economy from a micro level. Sohu News explores business operation strategies at a macro level. The Zhihu forum discusses the elements of digital development at the micro level. Moreover, the hot words’ degree centrality and betweenness centrality across various online communities exhibited a power law distribution. In conclusion, this research paper studies and analyzes DT themes of different social media platforms to discover the opinions and attitudes of various social groups in the digital transformation era and deeply interprets social trends and public opinions in order to provide valuable decision-making theoretical support for managers, enterprises, and governments. Full article

This study explores how linguistic and cultural differences shape social media discourses on green energy and sustainability by analysing English and Turkish tweets. Leveraging artificial intelligence-based text mining methods, the research examines users’ perceptions, emotions, and concerns about green energy on social media platforms. The findings reveal that in both languages, negative sentiments outweigh positive ones, with users frequently expressing their criticisms and apprehensions. However, significant thematic differences emerge based on language and culture. English tweets generally adopt a global and industrial perspective, while Turkish tweets are more focused on local, technical, and operational issues. By integrating sustainability into the analysis, this study highlights the interconnectedness of green energy discussions with broader environmental and societal goals. Social media platforms are shown to play a critical role in raising environmental awareness and influencing consumer perceptions. The results underline the importance of developing sustainability policies that consider regional dynamics, cultural contexts, and user expectations. Additionally, this study provides valuable insights for advancing climate research, media strategies, and digital marketing efforts. Ultimately, it emphasises the need for inclusive, informed, and innovative approaches to foster greener and more sustainable futures globally. Full article

Understanding user opinions from user comments or reviews in social media text mining is essential for marketing campaigns and many other applications. However, analyzing social media user comments presents significant challenges due to the complexity of discerning relationships between opinions and aspects, particularly when comments vary greatly in length. To effectively explore aspects and opinions in the sentences, techniques based on mining opinion sentiment of the referred aspects (implicitly or explicitly) in the user comments with ACOS (aspect-category-opinion-sentiment) quadruple extraction have been proposed. Among many others, the opinion tree parsing (OTP) scheme has been shown to be effective and efficient for the ACOS quadruple extraction task in aspect-based sentiment analysis (ABAS). In this study, we continue the efforts to design an efficient ABSA scheme. We extend the original OTP scheme further with richer context parsing rules, utilizing conjunctions and semantic modifiers to provide more context information in the sentence and thus effectively improving the accuracy of the analysis. Meanwhile, regarding the limitations of computation resources for edge devices in edge computing scenario, we also investigate the trade-off between computation saving (in terms of the percentage of model parameters to be updated) and the model’s performance (in terms of inference accuracy) on the proposed scheme under PEFT (parameter-efficient fine-tuning). We evaluate the proposed scheme on publicly available ACOS datasets. Experiment results show that the proposed enhanced OTP (eOTP) model improves the OTP scheme both in precision and recall measurements on the public ACOS datasets. Meanwhile, in the design trade-off evaluation for resource-constrained devices, the experiment results show that, in model training, eOTP requires very limited parameters (less than 1%) to be retrained by keeping most of the parameters frozen (not modified) in the fine-tuning process, at the cost of a slight performance drop (around 4%) in F1-score compared with the case of full fine-tuning. These demonstrate that the proposed scheme is efficient and feasible for resource-constrained scenarios such as for mobile edge/fog computing services. Full article