Feedback-type interactive features in electronic storybooks enhance learning regardless of cognitive differences.

RLBindDeep: A ResNet-LSTM based novel framework for protein-ligand binding affinity prediction.

Surface-water estradiol-benzo[a]pyrene mixtures: Transcriptomics in a PC12 neuronal differentiation model nominate an E2F1-PTPRO axis as a candidate mechanistic biomarker.

Protein-ligand binding affinity prediction is critical for drug development, where rapid and accurate assessment of drug-target binding remains a key challenge. Deep learning-based modeling provides a more efficient and scalable solution compared to traditional experimental approaches. This paper proposes a multi-scale interaction feature fusion model for protein-ligand binding affinity prediction, named the multi-scale binding affinity predictor (MSBind). The MSBind model employed multiple attention mechanisms to extract both global and local interaction features between the protein-ligand complex. By jointly modeling this multi-scale information, it ultimately enhanced the model's predictive performance. Experimental results showed that on the PDBbind v2016 core dataset, the root mean square error of MSBind was only 1.23, which was significantly lower than the baselines, demonstrating superior predictive accuracy. Furthermore, dimensionality reduction visualization of the extracted interaction features provided additional validation of the interpretability and effectiveness of MSBind. In summary, by integrating multi-scale interaction features, MSBind provides a high-performance solution for protein-ligand binding affinity prediction.

Finding the most suitable social media platforms in China for spreading musicals and related topics. Leveraging multiple resources to support and validate the research. Searching CNKI for literature related to social media and musicals. Designing and publishing a questionnaire on the frequency of social media use in China and the amount of reading on musical-related topics, to gain a timely and realistic understanding of the real-world situation. Interviewing relevant professionals and evaluate the suitability of each platform for musical-related topics from different perspectives. The literature demonstrates how different interactive features can be categorized, thereby identifying which questionnaires have the most valuable data. The survey results show that respondents prefer Rednote, and features with deeper interactive meaning are used more frequently on Rednote than on the other two platforms WeChat and Weibo. Due to its high interactivity and adaptability, Rednote is the most suitable social media platform for spreading musicals and related topics. This research can, on the one hand, promote the development of the musical and media industries, attracting more audiences, building positive reputation and value, and forming a virtuous cycle; on the other hand, it can promote musical culture, popularize niche culture, and enrich social and cultural values.

Purpose The study aims to explore the effect of destination digital twin (DT) experience on place attachment, tourists' intention to visit, and tourists' environmentally responsible behavior (TERB) in historical-cultural heritage sites. Design/methodology/approach This study utilizes a quantitative research approach. Data were gathered through online and offline surveys from 212 tourists who visited Naqsh-e Jahan, one of UNESCO's World Heritage Sites, through online and offline surveys. Furthermore, structural equation modeling (SEM) was employed for data analysis. Findings The findings show that DT experience dimensions (interactivity, informativeness, joyfulness and authenticity) are associated with place attachment, intention to visit and TERB. Additionally, place attachment positively and partially mediates the relationship between DT experience, visit intention and TERB. Practical implications The results indicate that heritage destination managers and digital tourism developers should design DT experiences that combine informative content, authentic representations, emotional engagement and interactive features in order to strengthen tourists' place attachment and encourage environmentally responsible behavior before on-site visitation. Originality/value This study contributes to enhancing digital experiences in historical-cultural heritage tourism by conceptualizing the multidimensional DT experience. The SEM findings support a model in which DT strengthens tourists' place attachment, both virtually and physically and confirm that place attachment partially mediates the relationship between DT experience (interactivity, informativeness, joyfulness and authenticity) and tourists' environmentally responsible behavior in sustainable historical-cultural tourism.

ABSTRACT In real industrial scenarios, distribution discrepancies in vibration signals across varying working conditions can weaken the generalisation capacity of fault diagnosis models, which negatively impacts diagnostic accuracy. To address this limitation, this paper proposes a multi-scale kernel invariant feature domain adversarial network (Msk-IFDAN). The method consists of three core components: an improved feature extractor, a classifier and a conditional domain discriminator. The feature extractor integrates a multi-scale kernel efficient channel attention (Msk-ECA) mechanism to achieve multi-scale channel interaction feature extraction and reduce feature redundancy. Meanwhile, invariant feature learning (IFL) is applied to learn domain-invariant features, enhancing the adaptability of the model to improve the generalisation capacity of cross-domain diagnosis of this method. The unified domain adaptation module of the design achieves precise alignment of feature distributions by reducing the distribution discrepancies through adversarial training strategies. Comprehensive comparative experiments on the Jiangnan University (JNU) dataset and Northeast Forestry University (NEFU) datasets demonstrate the effectiveness of the proposed method, with diagnostic accuracy reaching over 99%. The method not only enables accurate cross-domain fault diagnosis of rolling bearings but also exhibits good stability, effectively solving the challenges of cross-domain fault diagnosis for rolling bearings in industrial scenarios.

Acquiring a new language remains a significant challenge, particularly for slow learners who struggle with retention, comprehension, and motivation. With English recognized as a global lingua franca, there is an increasing demand for innovative pedagogical strategies that foster linguistic proficiency. This systematic review synthesizes evidence from 69 studies on Mobile-Assisted Language Learning (MALL), examining the effectiveness of mobile and gamified applications in supporting English language acquisition among slow learners. Drawing on the widespread accessibility of smartphones and the growing use of gamification, MALL applications employ adaptive content, interactive features, and varied theoretical frameworks to create personalized learning environments. These approaches address both cognitive limitations and motivational barriers by incorporating personalized pacing, iterative reinforcement, and real-time feedback. Contextual factors, including educational levels and learning settings, were found to play a critical role in shaping MALL outcomes. A novel taxonomy is proposed to classify existing MALL methodologies and technologies, providing a structured lens for future research and practical applications. Findings highlight the transformative potential of adaptive visuals and gamified strategies in enhancing learner engagement and language proficiency, while also acknowledging persistent challenges, including limited inclusivity, inconsistent design standards, and inadequate evaluation metrics. This review concludes that MALL, when designed with context-specific and learner-centered principles, has the potential to redefine English language education for slow learners. By leveraging mobile technologies and gamification, educators can create equitable, effective, and motivating pathways for learners who traditionally face barriers in language acquisition.

With the rapid development of the low-altitude economy, constructing an accurate unmanned aerial vehicle (UAV) air-to-ground channel model is crucial for ensuring communication quality. However, due to the significant fluctuations in UAV operation altitudes and the complex propagation environment, traditional empirical models struggle to achieve universal high-precision prediction within a 3D airspace. This paper proposes a Physics-Informed Feature Engineering (PIFE) method and constructs a 3D signal strength prediction model in combination with Gradient Boosting Decision Tree (XGBoost). Unlike traditional purely data-driven methods, this paper explicitly extracts physical propagation features such as three-dimensional Euclidean distance and height-to-angle ratio, and specifically designs a height–path loss interaction term to capture the nonlinear coupling relationship of signal attenuation at different operating heights. The experimental results demonstrate that the model proposed in this paper performs excellently in multi-altitude airspace scenarios ranging from 70 m to 150 m. At the typical operation height of 70 m, the model achieves a high goodness of fit (R2) of 0.843. Ablation experiments further confirm that the introduction of physical interaction features successfully breaks through the performance bottleneck of pure geometric features, proving the necessity of explicitly modeling the height–distance coupling effect in complex three-dimensional airspace. The research in this paper demonstrates the effectiveness of integrating physical priors with machine learning algorithms, providing an important theoretical basis and technical support for future drone network planning and coverage optimization in complex low-altitude environments.

This study examines the influence of the virtual try-on feature framed within the Unified Theory of Acceptance and Use of Technology (UTAUT) and online customer reviews on consumers' purchase intention toward Maybelline lipstick on Shopee. A quantitative approach was employed, with data collected from 114 active Shopee users in Parongpong District, West Bandung Regency. The data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) with SmartPLS software. The results indicate that effort expectancy, performance expectancy, and online customer reviews have a positive and significant influence on purchase intention in the context of using the virtual try-on feature on the Shopee platform. Of the three variables examined, online customer reviews were identified as the dominant factor influencing consumers' purchase intention. These findings indicate that integrating augmented reality technology via the virtual try-on feature, supported by a credible customer review system, can reduce consumer uncertainty and enhance consumer confidence when purchasing cosmetic products online. This study provides practical implications for e-commerce platforms and cosmetic companies, particularly in optimizing interactive digital features and strengthening customer review management systems as strategies to enhance consumers' purchase intention in online marketplaces.

ABSTRACT Considering the rapid growth of virtual reality (VR) applications in the industry, this study explores classifying virtual reality fashion shows (VRFS) from a user experience perspective. Prior research has not provided a systematic classification of VRFS, making it difficult for brands and consumers to evaluate different formats consistently. Key aspects examined include telepresence, immersion, interactivity, and technological features include model (real vs. virtual), scene environment (real vs. virtual), VR headset requirement, and platform types. Using data from academic journals, company websites, and metaverse-based platforms, a total of 140 VRFS videos were analyzed. Four types of VRFS were identified: fully immersive (Type 1), exclusive (Type 2), semi-immersive (Type 3), and naked-eye (Type 4). The findings provide insights for fashion companies to understand VRFS taxonomy and enhance consumer engagement through the strategic application of VR technology. This study contributes a replicable classification framework linking telepresence, immersion, and interactivity to distinct VRFS types.

This study investigates EFL students’ perceptions and learning experiences in using SmallTalk2Me, an AI powered speaking platform, to support speaking development. Employing a descriptive qualitative design, the study involved ten students from the English Education Study Program at the State Islamic University of Mataram, selected through purposive sampling. Data were collected through classroom observation during a mini workshop and semi structured interviews. The findings reveal that students generally perceived SmallTalk2Me positively because it provided immediate and useful feedback on fluency, pronunciation, grammar, and vocabulary. Contextualized prompts, interactive features, and CEFR based evaluations increased learners’ confidence, motivation, and willingness to speak more spontaneously. Participants also appreciated the platform’s natural conversational practice and its support for self directed learning. However, several challenges emerged, including limited access to free features, unstable internet connections, difficulty understanding some features, and a lack of topic variety, especially for beginners. Students suggested expanding conversation topics, providing more detailed corrective feedback, and making premium services more affordable. Overall, SmallTalk2Me shows strong potential as an adaptive medium for enhancing EFL speaking practice, although further development is needed to improve accessibility, inclusivity, and effectiveness in broader instructional contexts.

In this study, we researched how elementary and primary school students understand artificial intelligence in terms of its uses, mental representations, and the impact of gender, age, and digital experience. 521 students in Türkiye participated in the study, which was carried out as a basic qualitative investigation. Students’ responses were inductively coded in MAXQDA 24. According to the investigation, older students give more abstract and cognitively oriented explanations of AI, whilst the majority of kids link it to tangible technological instruments like computers, phones, and robots. Female students tend to express empathetic and anthropomorphic representations, whereas male students focus on technical and functional features. Higher degrees of digital device use were positively correlated with more complex conceptualizations of AI, such as interactive and human-like features. Overall, the results underscore the importance of AI-related designing of learning activities and pedagogical approaches that align with children’s developmental stages, gender differences, and digital experiences.

Advancements in information and communication technology have created new opportunities for developing accessible and adaptive learning systems. One promising approach is the use of audio-based learning applications, which substitute visual information with structured auditory instructions. This research aims to develop adaptive interactive features in the ‘Listen to Me” application to enhance the independent learning skills of visually impaired children. The features are designed with three main characteristics: automatic adaptation of material difficulty levels; an audio-based interface that delivers clear instructions and feedback; and dynamic responses that adjust to users’ interaction patterns. The study employed a Research and Development (R&D) approach with a waterfall model, executed through the stages of needs analysis, design, implementation, testing, and evaluation. Expert evaluations of the developed features indicated a very high level of feasibility. Furthermore, the results of limited trials involving the target users (visually impaired children) demonstrated the features’ effectiveness in improving content comprehension, facilitating navigation, and fostering motivation for independent learning. The findings conclude that the adaptive interactive features within the Listen to Me application are feasible, effective, and accessible as a self-learning medium, offering strong potential to support the expansion of inclusive education for visually impaired children.

Fragment-Based Drug Discovery (FBDD) is a powerful strategy with a proven track record of generating potent bioactive small molecules from low-affinity chemical fragments. Computational approaches to FBDD are often limited by the availability of high-quality, structurally resolved data on fragment binding poses. To address this gap, we introduce the Structurally Augmented Fragment Repository (SAFR), a novel data set designed to support in silico FBDD. Initially, a set of 89,375 high-confident binding poses of bioactive molecules in public sources was obtained by applying a filtering protocol involving 2D ligand similarity and 3D ligand superposition against protein-bound ligand structures followed by scoring with protein-ligand docking and interaction features. Fragmentation of the bioactive ligands in their predicted binding poses resulted in a total of 818,385 fragment-protein interactions between 157,080 unique chemical fragments and environments from 1,142 distinct proteins. Of them, 270,155 are unique fragment-protein interactions, of which 237,284 (88%) are not represented in protein-bound ligands in the PDB. Case studies using SAFR for bioisosteric replacements and scaffold hopping are presented. SAFR is a useful resource to support fragment screening campaigns and hit-to-lead optimization. It is publicly available at https://zenodo.org/records/18229523.

This research is motivated by obstacles in the implementation of character education at SMP Negeri 1 Loceret, particularly in the aspect of learning media that is less innovative in the era of digital transformation. The focus of this research is to develop Normapedia interactive media integrated with the SATRIA BAIK (One Day Three Goodness) program to strengthen students' moral actions in a real way. Using the Research and Development method with the ADDIE development model, this research goes through the stages of analysis, design, development, implementation, and comprehensive evaluation. The results of the validity test show a very high level of product feasibility reaching 91.87%, covering aspects of material, character, and media design. In the effectiveness test stage on 32 students, a significant increase in moral actions was found with an average score that jumped from 65.12 in the pretest to 84.56 in the posttest. This increase of 19.44 points is supported by an N-gain value of 0.71 which is in the high category and a significance of 0.000. Qualitatively, this program succeeded in increasing discipline consistency by 87% and social empathy by 81%. The main conclusion confirms that the synergy between interactive normative content and digital daily reflection features effectively builds a measurable and sustainable moral habitus. This innovation bridges the gap between students' understanding of ethical theory and their daily behavioral practices in a junior high school environment through the systematic use of technology. This medium supports the creation of a generation of intelligent and high-integrity learners.

Background Population aging has heightened concerns regarding social participation and active aging among older adults. Chatbot-based interventions delivered through familiar messaging platforms have increasingly been adopted to support engagement with community activity programs; however, prior research has often treated chatbots as homogeneous interventions, with limited attention to how specific functional components shape user experiences among older adults. Objective This study aimed to examine older adults’ functional experiences with a LINE-based chatbot designed to support engagement with community activity programs and to investigate how different chatbot functions are associated with overall system use experience. Methods A cross-sectional survey was conducted among 299 older adults who used a LINE-based chatbot implemented in community-based programs. Functional experiences were assessed across three chatbot modules: information inquiry, photo-based check-in interaction, and task achievement. Composite mean scores were calculated for each functional module and for the overall system use experience scale. Descriptive statistics, Pearson correlation analyses, and multiple linear regression analyses were performed using IBM SPSS Statistics. Results Participants reported moderate to moderately high levels of functional experience across all chatbot modules. Pearson correlation analyses showed significant positive associations among all functional experience variables. Multiple linear regression analysis indicated that the overall model explained a large proportion of variance in overall system use experience ( R 2 = 0.760, adjusted R 2 = 0.757, F (3,295) = 310.72, p < 0.001). Task achievement experience was the strongest predictor of overall system use experience ( β = 0.595, p < 0.001), followed by information inquiry experience ( β = 0.252, p < 0.001), whereas photo-based check-in experience was not significantly associated with overall system use experience ( β = 0.066, p = 0.139). Conclusion The findings suggest that different chatbot functions contribute unequally to older adults’ system use experience. Task-oriented and goal-focused functions may play a central role in shaping engagement with chatbot systems, whereas interactive features such as photo-based check-ins may serve a supplementary role. These results underscore the importance of function-specific design when developing chatbot-based interventions intended to support engagement with digital systems designed for community activity programs among older adults.

Three-dimensional (3D) in vitro models that more faithfully recapitulate human tissue architecture are essential to advance mechanistic insight into inflammatory diseases. We developed and characterized a 3D human gingival fibroblast spheroid-system to dissect the combined effects of hypoxia and inflammation, two central features of host-pathogen interactions in periodontal disease and other chronic inflammatory conditions. Using controlled oxygen environments and stimulation with E. coli lipopolysaccharide (LPS) and IL-1β, we show that hypoxia alone markedly alters spheroid morphology, producing smaller aggregates with thinner rims, while the combination of hypoxia and inflammation yields the most prominent structural disruption. Cross-section analyses reveal condition-dependent remodeling of extracellular matrix and cytoskeletal organization, with coordinated reductions in fibronectin, collagen I/III, integrin β1, and filamentous actin. Gene expression profiling confirms suppression of extracellular matrix (ECM)-related and adhesion genes under hypoxia and combined stress, whereas secretome analysis demonstrates robust upregulation of pro-inflammatory cytokines and chemokines (IL-6, IL-8, MCP-1, TNFα) and angiogenic VEGFα. Despite these marked structural and immunological shifts, cell viability remained largely preserved. This 3D system captures key features of fibroblast responses to inflammatory and hypoxic cues, providing a human-relevant platform for studies of tissue remodeling, cytokine signaling, and microenvironmental stress relevant to host-pathogen-driven disease.

Introduction Depression is a prevalent mental disorder with a severe global impact. Traditional interview-based assessments are limited by subjectivity, lengthy procedures, and unequal access to care. Although advances in AI have facilitated multimodal models for depression detection—using audiovisual data as an accessible alternative to biosignals—current approaches remain challenged by inefficient long-term temporal modeling and superficial multimodal fusion. Moreover, biosignal-based methods are constrained by high costs and narrow applicability. These challenges underscore the urgent need for optimized multimodal solutions. Methods This paper proposes ASYM (Attentive Synergy Mamba), a novel multimodal architecture for depression recognition, comprising three core modules: a Cross-Modal Interactive Mamba, a Multi-Scale Gated Parallel Fusion, and a Multimodal Enhanced Mamba. First, features from each modality are interactively enhanced using convolutional neural network and Bi-Mamba blocks. Cross-modal complementary information is then extracted via a cross-attention mechanism. A dual-path fusion module subsequently augments multi-scale representations and integrates cross-modal features through dynamic weighting. Finally, the feature representations are refined by a series of Bi-Mamba blocks. Results Evaluations on the D-Vlog and LMVD datasets using accuracy, precision, recall, and F1-score showed that ASYM achieved an accuracy of 70.91% and an F1-score of 77.13% on D-Vlog, and 74.68% accuracy with a 74.90% F1-score on LMVD. The macro-average performance across both datasets surpassed all compared mainstream methods. Ablation studies confirmed the necessity of each component, as removing any module significantly degraded performance, underscoring the efficacy and critical contribution of the proposed architecture. Discussion While multimodal depression detection has improved upon single-modality approaches, issues such as computational inefficiency in long-sequence processing and inadequate fusion strategies persist. Our model addresses these limitations through multimodal interaction and multi-scale feature fusion. Future work will focus on clinical validation across diverse populations to bridge computational psychiatry and clinical practice.

The increasing prevalence of paralysis and mobility-related disorders has created a strong demand for continuous healthcare monitoring systems, with neurological disorders affecting millions globally and nearly 60% of patients requiring constant assistance, while IoT-based healthcare solutions are projected to grow at over 20% annually. Traditional patient care relies heavily on manual observation and periodic monitoring, which may fail to detect sudden health deterioration, falls, or emergency conditions in time. Furthermore, conventional systems lack real-time alerts, remote accessibility, and interactive assistance features, reducing their effectiveness in supporting paralysis patients. To address these challenges, the proposed IoT-based health assistance system for paralysis patients utilizes the ESP32 microcontroller to develop an intelligent and wearable healthcare solution. The system integrates DHT11 sensors for temperature and humidity monitoring, SpO2 sensors for blood oxygen level measurement, vibration sensors for fall detection, and flex sensors to capture patient movements or specific gestures. When abnormal conditions or predefined triggers are detected, the system activates a voice module to alert nearby caregivers and simultaneously sends real-time notifications through an IoT platform for remote monitoring. This smart system enhances patient safety, enables rapid emergency response, improves caregiver support, and provides a scalable and efficient solution for modern assistive healthcare applications.