Search Results (4,655)

Sentiment analysis of online game reviews has increasingly become important in understanding player experiences and supporting data-driven game development. However, research in this domain has continuously faced two unresolved challenges: (1) the extreme imbalance between positive and negative feedback, and (2) the inefficiency of existing feature-weighting schemes in capturing sentiment signals embedded in informal gaming discourses. Prior works demonstrated that negative feedback—though a few in number are highly influential—usually contain richer emotional content and longer textual structures; yet, prevailing classification models often perform poorly for these minorities (i.e., negative feedback). Numerous studies explored multimodal imbalance issues, class imbalance in cross-lingual ABSA (Aspect-Based Sentiment Analysis), reinforcement-learning-based architectures for imbalanced extraction tasks, and oversampling strategies like SMOTE (Synthetic Minority Over-sampling Technique) variants. Few investigations specifically addressed imbalanced sentiment classification in the contexts of online game reviews, where user-generated content exhibits unique lexical, structural, and emotional characteristics. To address these gaps, this study integrated TF-IDF (Term Frequency-Inverse Document Frequency), VADER (Valence Aware Dictionary and Sentiment Reasoner) lexicon features, and IGM (Inverse Gravity Moment) weightings with advanced oversampling methods such as ADASYN (Adaptive Synthetic Sampling Approach for Imbalanced Learning) and Borderline-SMOTE to improve the detection of minority sentiment classes. Ensemble models, including XGBoost (Extreme Gradient Boosting) and LightGBM (Light Gradient-Boosting Machine), were further employed to enhance the robustness of imbalance. Using a large-scale dataset of Steam game reviews, the proposed framework demonstrated substantial improvement in identifying negative sentiments, addressing a critical limitation in the existing computational game-analysis literature, and advancing the modeling for detecting the emotion-rich but imbalance-prone user feedback. Full article

This study explores how school leadership styles are perceived to relate to teacher resilience during crises in Arab schools in Israel. Drawing on twenty semi-structured interviews with principals and vice-principals, findings show that transformational and participative leadership, characterized by emotional support, accessibility, active listening, and shared decision-making, are perceived to foster teachers’ sense of security, self-efficacy, and collective resilience. In contrast, authoritarian and rigid approaches are described as contributing to increased stress, reduced motivation, and diminished coping capacity. The study highlights the significance of socio-cultural and political contexts, indicating that effective leadership in crises involves not only professional guidance but also cultural awareness, flexibility, and responsiveness to staff needs. These findings underscore the value of integrative leadership approaches and targeted professional development to support teacher well-being and organizational resilience in crisis-prone settings. By focusing on leaders’ perspectives, the study contributes to understanding how culturally sensitive and adaptive leadership practices may support educational stability under conditions of uncertainty. Full article

Generative Artificial Intelligence (GenAI) is transforming higher education in terms of pedagogy, student involvement, and academic management. This systematic literature review examines 30 peer-reviewed articles published from 2019 to 2025, adhering to PRISMA 2020 and Kitchenham’s methodologies. Descriptive and thematic analyses highlight five opportunities: (a) tailored and adaptive education; (b) deliberate fostering of critical thinking; (c) enhanced accessibility for varied learners; (d) teaching innovation via multimodal content development and feedback; and (e) collaborative methods that regard AI as a co-teacher. Four ongoing challenge categories also surface: (a) risks to academic integrity; (b) excessive dependence on GenAI that may hinder learner independence; (c) inconsistent faculty preparedness and change-management abilities; and (d) differences in infrastructure and policy both regionally and globally. Intersecting ethical issues, such as data privacy, algorithmic bias, transparency, and accountability, highlight the necessity for governance that aligns with institutional risk and reflects societal values. Analyzing the recent literature, this systematic review offers four contributions: (a) a recommendation model for responsible GenAI implementation in higher education institutions; (b) a framework for sustainable integration of GenAI; (c) a highlight of the future research recommendations; and (d) an integrated policy and pedagogical recommendations roadmap. These models emphasize the integration of AI literacy, ethical considerations, and critical thinking goals into educational programs. The review advocates for a strategic, stakeholder-focused approach to implementation that enhances rather than replaces human instruction, thus connecting GenAI’s educational potential with ethical, context-aware avenues for institutional transformation. Full article

Accurate diagnosis of wheat leaf diseases in real farming conditions requires models that are not only highly accurate but also computationally efficient and interpretable for practical deployment on edge devices. We propose GateProtoNet (GPN), a two-stage, compute-aware, and explainable framework for multi-class leaf disease recognition. Stage-1 performs ultra-light healthy-versus-diseased screening, enabling early exit for healthy samples and substantially reducing average expected inference cost. For diseased samples, Stage-2 applies a novel hybrid backbone featuring a frequency-factorized Discrete Wavelet Transform (DWT) stem, parallel micro-lesion convolutional encoding for fine texture patterns, and a linear token mixer for global context modeling. A cross-gated fusion module adaptively integrates local and global evidence with minimal computational overhead. To ensure trustworthy predictions, GPN introduces a prototype evidence head that performs classification via similarity to learned class prototypes, providing human-interpretable explanations, along with a faithfulness constraint that enforces explanation reliability by measuring confidence degradation under salient region removal. Rigorous evaluation on four publicly available wheat and cotton leaf disease datasets demonstrate that GateProtoNet achieves 99.2% classification accuracy, 99.1% macro-F1 score, and 99.3% AUC, significantly outperforming existing CNN, transformer, and hybrid baselines while requiring substantially fewer parameters and FLOPs. The two-stage inference strategy reduces average computational cost by avoiding full model execution on healthy leaves, enabling real-time, on-device diagnosis for resource-constrained agricultural environments. Full article

This study investigates how CEO demographic characteristics, including age, gender, and nationality, and cognitive characteristics, including tenure, education, and multiple directorships, influence firms’ ESG performance, with a focus on the moderating role of Worldwide Governance Indicators (WGIs). Using a regime/smooth transition approach with panel data from STOXX Europe 600 firms spanning the years 1999 and 2023, the results show that demographic characteristics exert a more consistent effect than cognitive effects in the full sample and in non-sensitive industries. In sensitive industries, however, both demographic and cognitive CEO traits significantly affect ESG performance. Older and female CEOs enhance ESG performance under strong worldwide governance indicators (WGIs) in the full sample and sensitive industries, whereas foreign CEOs perform better under weaker worldwide governance conditions. In non-sensitive industries, the patterns for female and foreign CEOs are reversed. Cognitive traits such as tenure and multiple directorships show limited influence, while higher educational qualifications improve ESG outcomes under weak governance but reduce them under strong governance across all samples. Overall, the findings highlight the importance of aligning CEO characteristics with the institutional governance environment to enhance corporate sustainability performance. This study contributes by examining how CEO demographic and cognitive characteristics affect ESG performance under varying country-level governance conditions. It also highlights sectoral differences between sensitive and non-sensitive industries and, by using a nonlinear (PSTR) approach, uncovers regime-dependent effects with implications for governance-aware CEO selection and ESG strategy. This study extends upper echelons and institutional theories by showing that the effect of CEO characteristics on ESG performance depends on country governance quality, offering insights for boards and policymakers seeking to align leadership selection with governance contexts to strengthen sustainability and accountability. Full article

Previous research provides valuable insight into energy knowledge, attitudes, and behavior in the context of energy literacy. However, a research gap exists in dealing with a comprehensive understanding of complex relationships on energy-related attributes for adolescents. In this aspect, utilizing the framework of the Value–Belief–Norm (VBN) theory, this study highlighted these complex relationships for the selected adolescents as potential future energy consumers and future decision makers. Participants of the study were a total of 530 8th-grade students attending public schools located in a district of Istanbul, Türkiye. To test the hypothesized connections among the latent constructs, Structural Equation Modeling (SEM) was performed. This dataset demonstrates that altruistic values develop pro-environmental beliefs and awareness of consequences directly, but ascribed responsibility indirectly. Interestingly, biospheric and egoistic values showed no significant influence. In line with VBN theory, these students with more strong beliefs about human-nature interdependence develop an awareness of the consequences of their own actions and ascribed responsibility. The finalized model reveals that the relevant behaviors were linked to personal norms that could be positively explained by pro-environmental beliefs, awareness of consequences, and ascribed responsibility. These adolescents believing in the severity of global warming, while focusing on human well-being could be active in creating sustainable energy consumption patterns. This research provides valuable insight into strategies for promoting behavior aimed at reducing the persistent rise in energy consumption. Full article

This article explores Daniel Kahneman’s concept of system thinking from his book Thinking Fast and Slow (2013) in the context of contemporary spiritual and theological reflection. The question studied here is: What does the intentional use of emotions, dreams and intuition described by Daniel Kahneman as System 1 thinking look like in contemporary spiritual and theological reflection? According to Kanheman, System 1 thinking includes emotions, dreams and intuition. The method for answering the research question is hermeneutical. This means gathering texts that fit Kahneman’s description of System 1 thinking and integrating these concepts into some form of spiritual and theological reflection. Hermeneutical research is text-based. Fifty-three (53) texts were found in a search of various databases. These texts are analyzed noting the impact of System 1 thinking on spiritual and theological reflection. Findings include the following: First, there is a rise in the number of texts using System 1 thinking in spiritual and theological reflection. Second, disciplines outside of theology are practicing spiritual reflection as part of their spiritual care. Third, these non-theological disciplines are also using System 1 thinking in their spiritual reflections. Fourth, there is an awareness and utilization of diverse cultures and faith experiences in spiritual reflection. Fifth, these texts indicate the growth of the demographic of people who are spiritual but not religious and a connection to dreams, emotions and intuition in spiritual and theological reflection. Sixth, there is also a developing overlap between spiritual and theological reflection. Cautions and gaps in the textual analysis are noted as well as future applications. Full article

Aquaculture marks the transition from the simple activity of harvesting aquatic animal resources, carried out through the catching practices of fishing, to the farming of aquatic organisms in fresh, brackish and sea waters, carried out through human intervention aimed at increasing production. To date, research is proceeding towards expanding the range of species that can be farmed, improving the number and quality of products, and reducing the environmental impact of aquaculture activities; these efforts are supported by the improvement of our knowledge of the biology of the relevant species, the significant updating/upgrading of the rearing technologies, and the increasing awareness of the importance of water quality in optimising farming conditions. While necessarily dependent on market demand, aquaculture needs to fully leverage its environmental potential; and the relationship between aquaculture and the environment requires a system of production that combines eco-compatibility and eco-sustainability. Here, we report and analyse insights and perspectives in eco-sustainable aquaculture, spanning from sustainability and innovation processes in aquaculture to antibiotic control and aquaculture ecosystem services, in the context of the United Nations Sustainable Development Goals. Full article

Cross-view geo-localization (CVGL) aims to match ground-level images with geo-tagged aerial imagery for precise localization, but remains challenging due to severe viewpoint discrepancies, partial correspondence, and significant domain shifts across geographic regions. While existing methods achieve high accuracy within specific datasets, their generalization ability to unseen environments is limited. In this paper, we propose GenGeo, a unified framework that integrates vision foundation model representations with a matching-aware aggregation mechanism to address these challenges. Specifically, we leverage DINOv2 to extract semantically rich and transferable features, and revisit the SALAD aggregation module in the context of CVGL. By employing a shared clustering strategy, the proposed framework projects cross-view features into a unified assignment space, enabling implicit semantic alignment across views, while the dustbin mechanism effectively filters unmatched and non-informative regions arising from partial correspondence. Extensive experiments on three large-scale benchmarks (CVUSA, CVACT, and VIGOR) demonstrate that GenGeo achieves state-of-the-art performance in cross-dataset generalization and consistently improves robustness under severe domain shifts and spatial misalignment. Notably, our method outperforms the baseline by 14.65% in Top-1 Recall on the CVUSA-to-CVACT transfer task. These results highlight the effectiveness of combining foundation model representations with matching-aware aggregation, and suggest that enforcing semantic consistency in a shared assignment space is a promising direction for generalizable cross-view geo-localization. Full article

Large language models demonstrate strong in-context learning (ICL) capabilities, allowing them to perform diverse tasks without fine-tuning. In knowledge graph question answering (KGQA), natural language questions are translated into SPARQL queries. Existing ICL approaches mainly rely on semantic similarity, often neglecting structural features. To address this limitation, we developed a structure-aware example selection strategy that integrates both semantic and structural patterns by abstracting Resource Description Framework (RDF) triples. We compare four strategies: (1) fully random, (2) semantic similarity, (3) same-type random, and (4) same-type semantic similarity. Experiments on LC-QuAD 1.0 using FLAN-T5 show that in non-fine-tuned settings, structure-aware semantic selection achieves the best results, highlighting the importance of structural congruence, while after fine-tuning, differences between strategies converge but diversity and semantic relevance remain beneficial. These findings demonstrate the critical role of example quality in ICL and provide empirical insights for KGQA design. Full article

Mediterranean forests are increasingly exposed to climate-related risks, including large wildfires, prolonged droughts and rural abandonment, making sustainable forest management (SFM) a key element for climate adaptation and territorial resilience. However, despite its recognised importance, the social acceptance of SFM remains insufficiently understood, particularly in relation to how public perceptions are shaped by media narratives and information ecosystems. This study addresses this gap by analysing the relationship between media framing and social acceptance of SFM in a Mediterranean context. A mixed-methods approach was applied in the Valencian region (Spain), combining (i) a systematic analysis of conventional and digital media, (ii) a system mapping exercise to identify dominant narratives and communication dynamics, and (iii) a population survey (n = 1070) focused on perceptions of forests, climate change and forest management. The results reveal a high level of environmental concern and climate awareness, coexisting with limited knowledge of SFM and simplified or distorted perceptions of forest dynamics. Media coverage is predominantly reactive and event-driven, strongly focused on wildfire events, while preventive and adaptive forest management practices remain largely invisible. In this context, support for SFM increases significantly when management practices are clearly explained and contextualised, indicating that resistance is more closely related to communication gaps than to ideological opposition. These findings highlight the critical role of media framing and communication processes in shaping the social acceptance of SFM. The study contributes to the literature by integrating media analysis and social perception within a forest governance perspective, and provides empirical insights to support more effective communication strategies and policy design in Mediterranean regions facing increasing climate pressures. Full article

Sustainability and the Sustainable Development Goals (SDGs) are garnering significant attention due to growing global challenges, including poverty, inequality, environmental degradation, and climate change, with the latter addressed specifically through SDG 13. This study examined the level of self-reported awareness of six science-related SDGs—SDG 3 (Good Health and Well-Being), SDG 6 (Clean Water and Sanitation), SDG 7 (Affordable and Clean Energy), SDG 13 (Climate Action), SDG 14 (Life Below Water), and SDG 15 (Life on Land)—among science teachers in the Arab sector in Israel as a function of background variables: gender, seniority, degree type, academic institution, school type, area of specialization, and the integration of these SDGs into the science curriculum. The study employed a mixed-methods approach: in the quantitative component, 204 science teachers responded to a Likert-scale questionnaire; the qualitative component consisted of semi-structured interviews with 30 middle school science teachers from the Arab sector. The findings indicated a moderate level of self-assessed awareness regarding SDGs. Significant differences in awareness were found according to teaching subject: environmental studies teachers demonstrated the highest awareness, followed by general science, biology, and physics teachers, with chemistry teachers ranking lowest. No significant differences were found for the remaining variables (p > 0.05). Qualitative findings indicated that while teachers perceived SDG-related content as implicitly present in the curriculum, explicit and systematic integration of the SDG framework is largely absent. Overall, the findings suggest that teachers are not adequately exposed to the SDGs. Therefore, it is recommended to incorporate these topics into teacher-training courses and professional development programs and to further integrate them into curricula. This study contributes to the growing body of research on SDG integration in science education, particularly within underexplored minority educational contexts. Full article

Pedestrian gender recognition plays an important role in pedestrian analysis and intelligent video applications, for example, in demographic statistics, soft biometric analysis, and context-aware person retrieval. However, it remains a challenging task owing to viewpoint variations, illumination changes, occlusions, and low image quality in real-world imagery. To address these issues, an effective adaptive decision fusion framework, termed the Decision Fusion Learning Network (DFLN), is proposed in this paper. The key novel aspect of DFLN is that it effectively explores both an appearance-centered view that emphasizes detailed texture and clothing information and a structure-centered view that captures rich contour and structural information for pedestrian gender recognition. To realize DFLN, a Parallel CNN Prediction Probability Learning Module (PCNNM) is first constructed to independently learn modality-specific probabilities from color image and edge maps. Subsequently, a learnable Decision Fusion Module (DFM) is designed to fuse the modality-specific probabilities and explore their complementary merits for realizing accurate pedestrian gender recognition. The DFM can be easily coupled with the PCNNM, forming an end-to-end decision fusion learning framework that simultaneously learns the feature representations and carries out adaptive decision fusion. Experiments on two pedestrian benchmark datasets, named PETA and PA-100K, show that DFLN achieves competitive or superior performance compared with several state-of-the-art pedestrian gender recognition methods. Extensive experimental analysis further confirms the effectiveness of the proposed decision fusion strategy and its favorable generalization ability under domain shift. Full article

As in other cultures, university students in Nepal struggle with significant academic pressure, which often leads to academic anxiety and depression. The current study aims to expand awareness of the presence, prevalence, and impact of student academic anxiety and depression among Nepalese university students as well as to test an emerging approach to universal screening to identify learners’ need profiles to promote targeted intervention supports. Participants included 547 Nepalese college students who completed the Academic Anxiety Scale (AAS) and the University Student Depression Inventory (USDI). Confirmatory factor analysis (CFA) was conducted to evaluate the validity of the Nepalese versions. Finally, comparative analyses using an archival dataset of students from the United States explored consistencies across cultural contexts. Nepalese translations of both scales demonstrated high reliability and validity and identified similarities in patterns of expressed academic anxiety and depression across cultures. Furthermore, four profiles of need were generated based on levels of anxiety, depression, and academic motivation. The results supported clear recommendations for tiered interventions in specific domains of emotion regulation. This initial large-scale study of academic anxiety and depression in a Nepalese university population provided confirmation that the models of anxiety and depression as well as incidence levels were consistent with existing research from other contexts. Moreover, the results provided strong confirmation that universal screening with simplified self-report measures can identify clear patterns of need among students, which can be aligned with targeted tiered interventions to support student thriving. Full article

In the context of the rapid development of smart horticulture, a deep remote sensing-based dual detection method for horticultural crop growth anomalies and safety risks was proposed to address the limitations of existing remote sensing monitoring approaches. These conventional methods, which predominantly focused on growth vigor assessment or single-task anomaly detection, had difficulty distinguishing anomalies from actual production risks and exhibited insufficient sensitivity to weak anomalies and complex temporal disturbances. Within a unified framework, a growth state modeling branch and an anomaly perception branch were constructed, enabling the joint modeling of normal growth trajectories and anomalous deviation features. By further introducing a risk joint discrimination mechanism, an integrated analysis pipeline from anomaly identification to risk assessment was achieved. Multi-temporal remote sensing features were used as inputs, through which normal crop growth patterns were characterized via trend perception, texture modeling, and temporal aggregation, while sensitivity to local disturbances and weak anomaly signals was enhanced by anomaly embeddings and energy representations. Systematic experiments conducted on multi-regional and multi-crop horticultural remote sensing datasets demonstrated that the proposed method significantly outperformed comparative approaches, including traditional threshold-based methods, support vector machines, random forests, autoencoders, ConvLSTM, and temporal transformer models. In the dual task of horticultural crop growth anomaly detection and safety risk identification, an accuracy of approximately 0.91 and an F1 score of 0.88 were achieved, indicating higher anomaly recognition accuracy and more stable risk discrimination capability. Further anomaly-type awareness experiments showed that consistent performance was maintained across diverse real-world production scenarios, including climate stress, disease-induced anomalies, and management errors. Full article