Search Results (153)

Visual Impression in Architectural Space (VIAS) plays a central role in user response to environments, yet designer-controlled spatial variables often produce uncertain perceptual outcomes across cultural contexts. This study develops a multi-modal framework integrating VIAS theory, spatial documentation, and sentiment-aware NLP to evaluate temporary event spaces. Using a monthly market in Matsue, Japan as a case study, we introduce (1) systematic documentation of controlled spatial variables (layout, visibility, advertising strategy, (2) culturally balanced datasets comprising native Japanese and international participants across onsite, video, and virtual interviews, and (3) an adaptive sentiment-weighted keyword extraction algorithm suppressing interviewer bias and verbosity imbalance. Results demonstrate systematic modality effects: onsite participants exhibit festive atmosphere bias (+18% positive sentiment vs. video), while remote modalities elicit balanced critique of signage clarity and missing amenities. Cross-linguistic analysis reveals native participants emphasize holistic atmosphere, whereas international participants identify discrete focal points. The adaptive algorithm reduces verbosity-driven score inflation by 45%, enabling fair cross-participant comparison. By integrating spatial variable documentation with sentiment-weighted linguistic patterns, this framework provides a replicable methodology for validating architectural intent through computational analysis, offering evidence-based guidance for inclusive event space design. Full article

With the continuous evolution of social media, users are increasingly inclined to express their personal emotions on digital platforms by integrating information presented in multiple modalities. Within this context, research on image–text sentiment analysis has garnered significant attention. Prior research efforts have made notable progress by leveraging shared emotional concepts across visual and textual modalities. However, existing cross-modal sentiment analysis methods face two key challenges: Previous approaches often focus excessively on fusion, resulting in learned features that may not achieve emotional alignment; traditional fusion strategies are not optimized for sentiment tasks, leading to insufficient robustness in final sentiment discrimination. To address the aforementioned issues, this paper proposes a Dual-path Interaction Network with Multi-level Consistency Learning (DINMCL). It employs a multi-level feature representation module to decouple the global and local features of both text and image. These decoupled features are then fed into the Global Congruity Learning (GCL) and Local Crossing-Congruity Learning (LCL) modules, respectively. GCL models global semantic associations using Crossing Prompter, while LCL captures local consistency in fine-grained emotional cues across modalities through cross-modal attention mechanisms and adaptive prompt injection. Finally, a CLIP-based adaptive fusion layer integrates the multi-modal representations in a sentiment-oriented manner. Experiments on the MVSA_Single, MVSA_Multiple, and TumEmo datasets with baseline models such as CTMWA and CLMLF demonstrate that DINMCL significantly outperforms mainstream models in sentiment classification accuracy and F1-score and exhibits strong robustness when handling samples containing highly noisy symbols. Full article

With the rapid growth of visual content, automated aesthetic evaluation has become increasingly important. However, existing research faces three key challenges: (1) the absence of datasets combining Image Aesthetic Assessment (IAA) scores and Image Aesthetic Captioning (IAC) descriptions; (2) limited integration of quantitative scores and qualitative text, hindering comprehensive modeling; (3) the subjective nature of aesthetics, which complicates consistent fine-grained evaluation. To tackle these issues, we propose a unified multimodal framework. To address the lack of data, we develop the Textual Aesthetic Sentiment Labeling Pipeline (TASLP) for automatic annotation and construct the Reddit Multimodal Sentiment Dataset (RMSD) with paired IAA and IAC labels. To improve annotation integration, we introduce the Aesthetic Category Sentiment Analysis (ACSA) task, which models fine-grained aesthetic attributes across modalities. To handle subjectivity, we design two models—LAGA for IAA and ACSFM for IAC—that leverage ACSA features to enhance consistency and interpretability. Experiments on RMSD and public benchmarks show that our approach alleviates data limitations and delivers competitive performance, highlighting the effectiveness of fine-grained sentiment modeling and multimodal learning in aesthetic evaluation. Full article

Detecting fake news is essential in natural language processing to verify news authenticity and prevent misinformation-driven social, political, and economic disruptions targeting specific groups. A major challenge in multimodal fake news detection is effectively integrating textual and visual modalities, as semantic gaps and contextual variations between images and text complicate alignment, interpretation, and the detection of subtle or blatant inconsistencies. To enhance accuracy in fake news detection, this article introduces an ensemble-based framework that integrates textual and visual data using ViLBERT’s two-stream architecture, incorporates VADER sentiment analysis to detect emotional language, and uses Image–Text Contextual Similarity to identify mismatches between visual and textual elements. These features are processed through the Bi-GRU classifier, Transformer-XL, DistilBERT, and XLNet, combined via a stacked ensemble method with soft voting, culminating in a T5 metaclassifier that predicts the outcome for robustness. Results on the Fakeddit and Weibo benchmarking datasets show that our method outperforms state-of-the-art models, achieving up to 96% and 94% accuracy in fake news detection, respectively. This study highlights the necessity for advanced multimodal fake news detection systems to address the increasing complexity of misinformation and offers a promising solution. Full article

This paper presents the design, implementation and evaluation of an agentic virtual assistant (VA) for a medical clinic, combining large language models (LLMs) with retrieval-augmented generation (RAG) technology and multi-agent artificial intelligence (AI) frameworks to enhance reliability, clinical accuracy and explainability. The assistant has multiple functionalities and is built around an orchestrator architecture in which a central agent dynamically routes user queries to specialized tools for retrieval-augmented question answering (Q&A), document interpretation and appointment scheduling. The implementation combines LangChain and LangGraph with interactive visualizations to track reasoning steps, prompts using Gemini 2.5 Flash defines tool usage and strict formatting rules, maintaining reliability and mitigating hallucinations. Prompt engineering has an important role in the implementation and thus, it is designed to assist the patient in the human–computer interaction. Evaluation through qualitative and quantitative metrics, including ROUGE, BLEU, LLM-as-a-judge and sentiment analysis, confirmed that the multi-agent architecture enhances interpretability, accuracy and context-aware performance. Evaluation shows that the multi-agent architecture improves reliability, interpretability and alignment with medical requirements, supporting diverse clinical tasks. Furthermore, the evaluation shows that Gemini 2.5 Flash combined with clinic-specific RAG significantly improves response quality, grounding and coherence compared with earlier models. SBERT analyses confirm strong semantic alignment across configurations, while LLM-as-a-judge scores highlight the superior relevance and completeness of the 2.5 RAG setup. Although some limitations remain, the updated system provides a more reliable and context-aware solution for clinical question answering. Full article

To demonstrate the interrelation of fashion elements for design forecasting, the research examines classification and sentiment analysis methodologies. The study combines survey data with information from social media and e-commerce sites to find important emotional and behavioral patterns that affect how people make buying decisions. The research employs deep learning, Logistic Regression, and Random Forest models to predict design trends and user preferences. The research methodology focuses on improving fashion analytics through feature selection and user segmentation and visual storytelling methods to enhance strategic decision-making. Full article

With the rapid development of mobile internet technology, the explosive growth of image–text multimodal data generated by social networking platforms has provided rich practical scenarios and theoretical research value for multimodal sentiment analysis. However, existing methods generally suffer from inefficient modal interaction and imperfect sentiment aggregation mechanisms, particularly an over-reliance on visual modalities, leading to an imbalance in cross-modal semantic correlation modeling. To address these issues, this paper proposes a sentiment analysis algorithm for image–text modalities based on curriculum learning and attention mechanisms. The algorithm introduces the concept of curriculum learning, fully considering the negative impact of irrelevant images in image–text data on overall sentiment analysis, effectively suppressing interference from irrelevant visual information without requiring manual data cleaning. Meanwhile, the algorithm designs a dual-stage attention architecture—first capturing cross-modal correlation features via cross-modal attention, then introducing an attention bottleneck strategy to compress redundant information flow, achieving efficient feature fusion by constraining intra-modal attention dimensions. Finally, extensive experiments were conducted on two public datasets, demonstrating that the proposed method outperforms existing approaches in sentiment prediction performance. Full article

Multimodal sentiment analysis jointly exploits textual, acoustic, and visual signals to recognize human emotions more accurately than unimodal models. However, real-world data often contain noisy or partially missing modalities, and naive fusion may allow unreliable signals to degrade overall performance. To address this, we propose an enhanced framework named MISA-GMC, a lightweight extension of the widely used MISA backbone that explicitly accounts for modality reliability. The core idea is to adaptively reweight modalities at the sample level while regularizing cross-modal representations during training. Specifically, a reliability-aware gated fusion module down-weights unreliable modalities, and two auxiliary training-time regularizers (momentum contrastive learning and a lightweight correlation graph) help stabilize and refine multimodal representations without adding inference-time overhead. Experiments on three benchmark datasets—CMU-MOSI, CMU-MOSEI, and CH-SIMS—demonstrate the effectiveness of MISA-GMC. For instance, on CMU-MOSI, the proposed model improves 7-class accuracy from 43.29 to 45.92, reduces the mean absolute error (MAE) from 0.785 to 0.712, and increases the Pearson correlation coefficient (Corr) from 0.764 to 0.795. This indicates more accurate fine-grained sentiment prediction and better sentiment-intensity estimation. On CMU-MOSEI and CH-SIMS, MISA-GMC also achieves consistent gains over MISA and strong baselines such as LMF, ALMT, and MMIM across both classification and regression metrics. Ablation studies and missing-modality experiments further verify the contribution of each component and the robustness of MISA-GMC under partial-modality settings. Full article

This paper presents a systematic literature review (SLR) of research on strategic positioning of companies and the measurement of greenwashing in sustainability reporting. Its main aim is to synthesize and organize the existing literature, identify key research gaps, and outline directions for future studies. Drawing on a rigorous content analysis of 88 studies, we delineate strategic typologies of greenwashing in sustainability reporting and examine content-analysis-based measurement approaches used to detect it. Our SLR shows that most strategic typologies draw on theories such as legitimacy theory, impression management theory, signaling theory, and stakeholder theory. Several studies adopt a four-quadrant matrix with varying conceptual dimensions, while others classify strategic responses to institutional pressures along a passive–active continuum. However, the evidence suggests that to assume that companies uniformly pursue sustainability reporting strategies is a major oversimplification. The findings also indicate that the literature proposes a variety of innovative, content-analysis-based approaches aimed at capturing divergences between communicative claims and organizational realities—most notably, discrepancies between disclosure and measurable performance, and between symbolic and substantive sustainability actions, as well as the identification of selective or manipulative communication practices that may signal greenwashing. Analytical techniques commonly focus on linguistic and visual cues in sustainability reports, including tone (sentiment and narrative framing), readability (both traditional readability indices and machine learning–based textual complexity measures), and visual content (selective emphasis, imagery framing, and graphic distortions). We also synthesize studies that document empirically verified instances of greenwashing and contrast them with research that, in our view, relies on overly simplified or untested assumptions. Based on this SLR, we identify central theoretical and methodological priorities for advancing the study of greenwashing in sustainability reporting and propose a research agenda to guide future research. Full article

Non-Fungible Tokens (NFTs) have transformed digital ownership, offering unique representations of assets such as art, collectibles, and virtual property. However, pricing NFTs remains a complex and underexplored issue. This study addresses two core questions: what determines NFT prices? And how are prices set in NFT markets? We conduct a comprehensive literature review and market analysis to identify both endogenous and exogenous price determinants. Trait rarity emerges as the most influential intrinsic factor, while cryptocurrency value stands out as a major external influence, albeit with ambiguous effects. Other factors include visual aesthetics, scarcity, utility in games, social media engagement, and broader market sentiment. As to pricing mechanisms, aside from fixed pricing (which is accepted in all marketplaces), NFT marketplaces primarily utilise auctions for art pieces and collectibles— especially English and Dutch formats—which are effective at capturing the buyer’s willingness-to-pay. Full article

Using Shanghai as a case study, this paper develops a multi-source fusion and interpretable machine learning framework. Sentiment indices were extracted from Weibo check-ins with ERNIE 3.0, street-view elements were identified using Mask2Former, and urban indicators like the Normalized Difference Vegetation Index, floor area ratio, and road network density were integrated. The coupling between residents’ sentiments and streetscape features during heatwaves was analyzed with Extreme Gradient Boosting, SHapley Additive exPlanations, and GeoSHAPLEY. Results show that (1) the average sentiment index is 0.583, indicating a generally positive tendency, with sentiments clustered spatially, and negative patches in central areas, while positive sentiments are concentrated in waterfronts and green zones. (2) SHapley Additive exPlanations analysis identifies NDVI (0.024), visual entropy (0.022), FAR (0.021), road network density (0.020), and aquatic rate (0.020) as key factors. Partial dependence results show that NDVI enhances sentiment at low-to-medium ranges but declines at higher levels; aquatic rate improves sentiment at 0.08–0.10; openness above 0.32 improves sentiment; and both visual entropy and color complexity show a U-shaped relationship. (3) GeoSHAPLEY shows pronounced spatial heterogeneity: waterfronts and the southwestern corridor have positive effects from water–green resources; high FAR and paved surfaces in the urban area exert negative influences; and orderly interfaces in the vitality corridor generate positive impacts. Overall, moderate greenery, visible water, openness, medium-density road networks, and orderly visual patterns mitigate negative sentiments during heatwaves, while excessive density and hard surfaces intensify stress. Based on these findings, this study proposes strategies: reducing density and impervious surfaces in the urban area, enhancing greenery and quality in waterfront and peripheral areas, and optimizing urban–rural interfaces. These insights support heat-adaptive and sustainable street design and spatial governance. Full article

Against the backdrop of Shenzhen’s high-density urban environment, the multifunctional design of water purification plants offers dual benefits: providing residents with urban green spaces while simultaneously mitigating NIMBY sentiments due to their inherent characteristics. Unlike traditional urban development, Shenzhen’s water purification plants integrate into residents’ daily lives. Therefore, optimizing the built environment and road network structure to enhance residents’ perceptions of proximity benefits while reducing NIMBY (Not In My Backyard effect) sentiments holds significant implications for the city’s sustainable development. To address this question, this study adopted the following three-step mixed-methods approach: (1) It examined the relationships among residents’ YIMBY (Neighboring Benefits Effect) and NIMBY perceptions, perceptions of park spaces atop water purification plants, and perceptions of accessibility through questionnaire surveys and structural equation modeling (SEM), establishing a scoring framework for comprehensive YIMBY and NIMBY perceptions. (2) Random forest models and Shapley Additive Explanations (SHAP) analysis revealed nonlinear relationships between the built environment and composite YIMBY and NIMBY perceptions. (3) Spatial syntax analysis categorized the upgraded road network around the water purification plant into grid-type, radial-type, and fragmented-type structures. Scatter plot fitting methods uncovered relationships between these road network types and resident perceptions. Finally, negative perceptions were mitigated by optimizing path enclosure and reducing visual obstructions around the water purification plant. Enhancing neighborhood benefits—through improved path safety and comfort, increased green spaces and resting areas, optimized path networks, and diversified travel options—optimized the built environment. This approach proposes design strategies to minimize NIMBY perceptions and maximize YIMBY perceptions. Full article

This paper proposes a hybrid deep learning model for robust and interpretable sentiment classification of Twitter data. The model integrates Bidirectional Encoder Representations from Transformers (BERT)-based contextual embeddings, a Bidirectional Long Short-Term Memory (BiLSTM) network, and a custom attention mechanism to classify tweets into four sentiment categories: Positive, Negative, Neutral, and Irrelevant. Addressing the challenges of noisy and multilingual social media content, the model incorporates a comprehensive preprocessing pipeline and data augmentation strategies including back-translation and synonym replacement. An ablation study demonstrates that combining BERT with BiLSTM improves the model’s sensitivity to sequence dependencies, while the attention mechanism enhances both classification accuracy and interpretability. Empirical results show that the proposed model outperforms BERT-only and BERT+BiLSTM baselines, achieving F1-scores (F1) above 0.94 across all sentiment classes. Attention weight visualizations further reveal the model’s ability to focus on sentiment-bearing tokens, providing transparency in decision-making. The proposed framework is well-suited for deployment in real-time sentiment monitoring systems and offers a scalable solution for multilingual and multi-class sentiment analysis in dynamic social media environments. We also include a focused characterization of the dataset via an Exploratory Data Analysis in the Methods section. Full article

As tangible relics of modern colonial history, Manchurian regime (Manchukuo) architecture of Changchun possesses both historical commemorative value and tourism and cultural functions. Public perception and sentiment regarding this heritage in the contemporary social media context are key dimensions for evaluating the effectiveness of cultural regeneration. Existing research on Manchurian regime architecture has focused primarily on historical research and architectural form analysis, with limited research examining the diverse public interpretations of its cultural value through multi-platform social media data. This study aims to systematically explore the public’s cognitive characteristics, sentimental attitudes, and themes of interest regarding Changchun’s Manchurian regime architecture using online ethnographic data, providing empirical support for optimizing cultural regeneration pathways for Manchurian regime architectural heritage. The study collected data from 1 January 2020 to 20 September 2025, using the keyword “Changchun Manchurian regime architecture”. Using Python crawlers, the study extracted 334 original videos and 18,156 related comments from Douyin, Ctrip, and Dianping. The analysis was conducted using word frequency statistics, SnowNLP sentiment analysis, LDA topic modeling, and multidimensional visualization. The study found that (1) word frequency statistics show that the public has multiple concerns about the historical symbols, geographical positioning, cultural and tourism functions, and national emotions of Manchurian regime architecture; (2) SnowNLP analysis shows that positive comments account for 71%, neutral comments account for 11%, and negative comments account for 18%; (3) the optimal number of topics was determined to be five through perplexity and consistency indicators, namely “historical narrative and imperial power symbols”, “emotional experience and historical reflection”, “visit experience and service facilities”, “site distribution and regional space”, and “explanation and tour evaluation”; (4) the corpus can be divided into five time period stages, namely S1 (2020)–S5 (2024–2025), reflecting the shift in public attention from “space-facilities” to in-depth reflection on “emotion-history”. Full article

To quantify the communication effectiveness of automotive online videos, this study constructs a multimodal deep learning framework. Existing research often overlooks the intrinsic and interactive impact of textual and dynamic visual content. To bridge this gap, our framework conducts an integrated analysis of both the textual (titles) and visual (frames) dimensions of videos. For visual analysis, we introduce FER-MA-YOLO, a novel facial expression recognition model tailored to the demands of computational communication research. Enhanced with a Dense Growth Feature Fusion (DGF) module and a multiscale Dilated Attention Module (MDAM), it enables accurate quantification of on-screen emotional dynamics, which is essential for testing our hypotheses on user engagement. For textual analysis, we employ a BERT model to quantify the sentiment intensity of video titles. Applying this framework to 968 videos from the Bilibili platform, our regression analysis—which modeled four distinct engagement dimensions (reach, support, discussion, and interaction) separately, in addition to a composite effectiveness score—reveals several key insights: emotionally charged titles significantly boost user interaction; visually, the on-screen proportion of human elements positively predicts engagement, while excessively high visual information entropy weakens it. Furthermore, neutral expressions boost view counts, and happy expressions drive interaction. This study offers a multimodal computational framework that integrates textual and visual analysis and provides empirical, data-driven insights for optimizing automotive video content strategies, contributing to the growing application of computational methods in communication research. Full article