[Purpose/Significance] The rapid development of embodied intelligence has fundamentally transformed the landscape of privacy protection. Unlike traditional digital services, embodied intelligent devices deeply embed themselves into users' physical environments, continuously collecting multimodal data through sensors and autonomously executing actions. This physical embodiment, social interaction, and action autonomy fundamentally reshape privacy risk boundaries, rendering existing privacy policy evaluation frameworks inadequate. While privacy policy evaluation has been extensively studied for websites and mobile applications, research specifically addressing the unique challenges of embodied intelligence remains scarce. This study aims to fill this gap by developing a user-friendliness evaluation framework tailored for embodied intelligence privacy policies. The theoretical innovation lies in extending the friendliness concept from traditional digital spaces to physical interaction scenarios, systematically incorporating the technical characteristics of embodied intelligence into the evaluation system. The practical significance is to reveal the current state of privacy policy design in the embodied intelligence industry, identify compliance gaps and interaction design deficiencies, and provide actionable recommendations for policy optimization and user rights protection. [Method/Process] This study constructed a comprehensive evaluation system grounded in user experience theory and legal requirements. Theoretically, we introduced Peter Morville's user experience honeycomb model as the foundational framework, which comprises seven interrelated dimensions: useful, usable, desirable, findable, accessible, credible, and valuable. This model provides systematic coverage of the entire user journey from initial perception to trust establishment. Legally, the indicator system refers to the personal information protection law and information security technology - personal information security specification. Methodologically, we innovatively adopt the matter-element extension model as the core evaluation approach. Compared to traditional methods like analytic hierarchy process and fuzzy comprehensive evaluation, this model offers unique advantages in handling multi-indicator, multi-grade, and subjective perception data, providing precise association degrees between each indicator and evaluation grades. Empirically, we collected 336 valid questionnaires from users with experience using embodied intelligence products, consulting five experts (three in user information behavior and two in human-computer interaction) to establish evaluation grade standards. Six mainstream embodied intelligence applications across different product categories were selected as evaluation objects, including DJI (autonomous drones), Ecovacs (sweeping robots), and NIO (in-vehicle intelligent systems). [Results/Conclusions] The evaluation results reveal three key findings. First, the overall user-friendliness of current embodied intelligence privacy policies ranges from moderate to good, exhibiting a middle-clustered distribution with no applications achieving excellent ratings, indicating the industry is transitioning from basic compliance to experience optimization. Second, regarding core obligation indicators, basic compliance is generally achieved, yet significant deficiencies exist in deepening user rights (e.g., inclusive protection for special groups) and transparent risk communication (e.g., specific disclosure of physical security risks). Notably, Yushu Technology and Qianglang Smart received poor ratings for user consent respect, as their interfaces lack substantive refusal options. Third, interaction friendliness indicators show uneven performance: DJI and Ecovacs excel in built-in functionality (achieving excellent ratings by integrating policy reading and settings within the app), while most applications underperform in visual design, language simplicity, and navigation convenience. This study validates the applicability of the matter-element extension model in privacy policy evaluation and provides actionable recommendations: optimize consent mechanisms to ensure clear refusal options, enhance visual hierarchy and layout design to reduce cognitive load, and specifically address unique embodied intelligence risks such as physical security and environmental monitoring. Future research should expand sample sizes, incorporate multimodal sensing data, and explore dynamic real-time evaluation mechanisms.

[Purpose/Significance] Against the backdrop of building a culturally strong nation and the requirements of the 15th Five-Year Plan for cultural digitalization and intelligent empowerment, the digital revitalization of cultural heritage has reached a critical juncture. It is transitioning from documentation and archiving to providing immersive experiences and reinventing value. Embodied intelligence, which emphasizes the dynamic interaction between "body, environment and cognition", can give cultural heritage tangible and interactive capabilities, thereby encouraging users to shift from passive perception to active participation. However, existing research lacks a systematic exposition of the intrinsic logic linking embodied intelligence and the digital revitalization of cultural heritage. This paper is the first to construct a theoretical framework and empowerment mechanism from a multi-layered, progressive perspective encompassing perception, interaction and cognition, whilst revealing the gradations of difference across various scenarios. It holds both academic value and practical significance for refining digital humanities theory, guiding practice and promoting the creative transformation of traditional culture. [Method/Process] A combined approach of theoretical analysis and case studies was adopted. The theoretical basis draws on the interactive nature of embodied intelligence, - specifically, the "body-environment-cognition" triad - and the digital revitalization pathway of cultural heritage, which involves transitioning from a "static resource layer" to a "dynamic narrative layer". An empowerment mechanism model was constructed across three levels: perception enhancement, interaction deepening, and cognition generation. Building upon this, three typical scenarios - virtual presence, craft simulation and project participation - were identified and analyzed. The case studies of "Seeking the Spirit of Dunhuang", the VR ceramic firing simulation system and the "Recognizing Classical Texts" platform were selected for examination, with comparisons made across dimensions such as core elements, technological dependencies and modes of cultural transmission. Through a synthesis of case studies and industry research, the study identifies current practical challenges, including high technical costs and limited accessibility, the potential for distortion in cultural expression and blurred ethical boundaries, as well as a lack of sustainable operation and maintenance alongside a one-dimensional evaluation mechanism. [Results/Conclusions] The study demonstrates that embodied intelligence holds immense potential for the digital revitalization of cultural heritage. Virtual presence overcomes temporal and spatial constraints to achieve an "immersive"experience; craft simulation transforms tacit knowledge into learnable bodily experience; and project participation facilitates large-scale social collaboration and co-creation of knowledge. The three types of scenarios exhibit a spectrum-like distribution in terms of the degree of bodily participation, technological dependence and modes of cultural transmission, collectively forming a complete chain of cultural transmission. To address these practical challenges, four development strategies are proposed. First, to establish a lightweight, embodied interaction model that is inclusive and accessible. Second, to establish comprehensive cultural and ethical safeguards covering the entire process from data collection and content generation to user interaction. Third, to improve sustainable operation and maintenance alongside a diversified evaluation system. Fourth, to build a collaborative network for the coordinated development of talent, technology and heritage. The limitations of this study include a focus on domestic case studies and a lack of large-scale user validation. Future research could explore cross-cultural comparisons and long-term impact assessments.

[Purpose/Significance] Generative AI has revitalized short video creation but blurred the line between authentic and false information. AI disclosure - informing the public of AI involvement via labels - has become a normative tool to mitigate ethical risks. However, current disclosure practices remain overly simplistic and fail to meet users' needs for transparency. How detailed disclosure labels should be designed and how they affect user cognition and behavior is still unclear. Previous studies have largely focused on the impact of whether AI is disclosed or not. Drawing on transparency design, this study further investigates the level of detail in AI disclosure. Theoretically, it confirms the importance of detailed disclosure, reveals the underlying mechanism via the heuristic-systematic model (HSM), and identifies the boundary condition of video type. Practically, it provides guidance for creators in designing disclosure labels and for platforms in effective regulation. [Method/Process] To examine the impact of AI disclosure detail on users' engagement intention, the study adopted an online experiment employing a 2×2 between-subjects experimental design to compare the effects of two types of disclosure labels across different video types. Detailed AI disclosure labels provided information regarding the type of AI technology used, its purpose, and its limitations, whereas simple AI disclosure labels merely indicated the involvement of AI technology. Additionally, the study distinguished between utilitarian and hedonic short videos for further investigation. Participants were recruited through an online experimental platform and randomly assigned to different experimental conditions. In each condition, participants were asked to view AI-generated short videos accompanied by the corresponding disclosure labels and completed a standardized questionnaire. Specifically, the questionnaire assessed participants' perceived source credibility, perceived content quality, and engagement intention, with measurement instruments adapted from validated scales. [Results/Conclusions] The study found that the level of AI disclosure detail has a significant positive effect on users' engagement intention with AI-generated short videos. Detailed AI disclosure enhances perceived content quality through the systematic route, thereby positively influencing users' engagement intention. However, the mediating role of perceived source credibility as a heuristic cue was not significant. Furthermore, this effect showed no significant difference across different types of short videos. The study provides valuable insights into the mechanisms of AI transparency. Future research can consider other dimensions of effective disclosure, explore the effects of dynamic prompts, expand sample coverage to enhance the applicability of conclusions, and conduct empirical tests incorporating actual behavioral data, thereby deepening the understanding of the mechanisms underlying AI disclosure effects.

[Purpose/Significance] To address the issues of knowledge dispersion, redundancy, and fragmentation within the organization of complex historical events, this study employs the event evolution graph method. It explores their potential applications in the semantic organization of historical events and the structured representation of historical knowledge. The aim is to enhance the discoverability, reusability, and semantic relevance of historical data, while broadening the theoretical framework and practical approaches of the event evolution graph in historical research. [Method/Process] First, we constructed a three-layer framework for complex historical event, consisting of a data layer, a semantic layer, and an application layer. The data layer enables the collection of all event elements and supports multimodal integration, including text, images, and maps. The semantic layer implements AI-enhanced event representation and deep relationship mining. The application layer performs correlation calculation and multi-level graph visualization, allowing users to interactively explore event structures and semantic pathways. On this basis, three key tasks were carried out. First, AI-enhanced methods for representing and extracting complex historical events were designed. In particular, a dynamic ontology model for AI understanding was constructed based on the W7 model, formally represented as:e = {what, why, who, how, which, when, where, environment, effect, certainty}. This formalization systematically depicts event elements and their semantic relationships, capturing not only basic event components but also contextual factors, causal consequences, and the degree of historical certainty. Second, AI-enhanced methods for calculating the correlation of complex historical events were proposed. These methods combine rule-based reasoning with machine learning classifiers to identify and quantify semantic relationships such as causality, temporality, correlation, and hierarchy among events. A case study on a representative complex historical event was conducted to validate the proposed framework and methods. [Results/Conclusions] The study demonstrates that AI can improve the accuracy of extracting event elements and analyzing semantic relationships. This provides a feasible technical pathway for organizing historical knowledge computationally and providing intelligent services. The case study results show that the generated event evolution graph captures multi-level event structures and reveals previously implicit causal and evolutionary patterns. However, research in this field still faces challenges, including the scarcity of high-quality historical corpora, subjectivity when generalizing and decomposing events, and insufficient integration of multimodal information. In the future, we will focus on three directions: developing weakly supervised learning and transfer learning methods tailored to scenarios with sparse historical data; designing human-computer collaborative tools for event decomposition and relationship annotation to balance automation with scholarly interpretability; and constructing multimodal event evolution graph for complex historical events by incorporating visual, spatial, and audio data.

[Purpose/Significance] This study aims to reveal the influencing factors that affect users' behavioral intention to adopt Artificial Intelligence (AI) smart services in public libraries. As public cultural institutions transition toward intelligent service paradigms, the integration of generative AI offers unprecedented opportunities to enhance knowledge accessibility and operational efficiency. By exploring users' actual needs for AI-driven tools - such as intelligent reference desks, personalized reading recommendations, and automated retrieval systems - this research seeks to provide robust theoretical and practical guidance. Ultimately, it aims to promote the deep integration of AI technologies within the broader framework of smart library construction, ensuring that these innovations align with user expectations and the public interest. [Method/Process] Drawing upon the Technology Acceptance Model (TAM) as the foundational theoretical framework, this study introduces Trust and Perceived Risk as critical external variables to accurately reflect the current technological climate, which is increasingly characterized by data privacy concerns and algorithmic opacity. Data were collected through a structured online questionnaire survey targeting a diverse demographic of public library users, resulting in 257 valid responses. To empirically test the proposed research model and hypotheses, Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed. The rigorous analytical process included a comprehensive assessment of the measurement model to confirm internal consistency, convergent validity, and discriminant validity, followed by the evaluation of the structural model to determine the statistical significance of the path coefficients and the overall explanatory power of the integrated framework. [Results/Conclusions] The empirical evaluation of the structural model yielded several key findings. First, both Perceived Usefulness (PU) and Perceived Ease of Use (PEOU) exert a significant positive impact on user satisfaction, highlighting that functional utility and intuitive interfaces are baseline requirements for AI adoption. Second, Trust, Satisfaction, PU, and PEOU are all identified as strong, direct positive predictors of users' Behavioral Intention (BI) to use AI smart services. Third, Perceived Risk (PR) significantly and negatively influences BI, acting as a major barrier to adoption. Interestingly, the influence of PR on PU was found to be statistically insignificant, suggesting that users evaluate the functional benefits of AI independently of its potential risks. Finally, Trust was shown to effectively mitigate user concerns, exerting a significant negative impact on PR. Based on these insights, it is recommended that public libraries prioritize enhancing the algorithmic transparency of their AI applications to systematically build user trust. Furthermore, libraries should integrate regional cultural elements to develop localized and distinctive AI services, diversify AI application scenarios to meet multifaceted user demands, and actively implement educational workshops and lectures focused on improving public AI literacy.

[Purpose/Significance] University libraries are experiencing a structural change in reference and consultation services: user inquiries are increasingly frequent, fragmented, and cross-disciplinary, while service expectations emphasize immediacy, continuity, and actionable guidance. Under such conditions, large language models may relieve routine workloads and extend service availability, yet their library-specific reliability hinges on whether their responses are grounded in local rules, licensed resources, and auditable evidence. This study examines the deployment of DeepSeek in intelligent QA and service consultation in university libraries, with two goals: 1) to measure its performance across consultation scenarios, disciplinary domains, and inquiry types; and 2) to clarify the mechanisms that explain why effectiveness improves in some settings but not in others. The study differentiates itself from prior discussions by moving beyond general "potential and risk" arguments to a structured evaluation and mechanism-oriented analysis that links local knowledge governance, scenario engineering, human - machine collaboration, and operational constraints to observable service outcomes. [Method/Process] The research adopts a mixed-method design and investigates five university libraries in Henan Province that have piloted or deployed DeepSeek-related services to varying degrees. Data sources include a) user questionnaires capturing usage frequency, task preferences, satisfaction, and perceived value; b) staff questionnaires documenting deployment modes, knowledge-base connections, maintenance routines, quality control practices, and operational challenges; c) in-depth interviews that detail workflow design, escalation rules, and the division of labor between librarians and the system; and d) case materials and system records used for triangulation. In total, 850 questionnaires were distributed and 783 valid responses were collected (625 users and 158 staff). An evaluation framework was constructed along four dimensions - technical performance, service effectiveness, user experience, and managerial benefits - to ensure comparability across libraries. Quantitative analyses include descriptive statistics and group comparisons across inquiry types and disciplines, supplemented by mechanism-oriented interpretation using indicators such as the depth of local knowledge integration, the effectiveness of retrieval augmentation, the degree of scenario customization, and the intensity of governance constraints. Qualitative coding of interviews and case materials was conducted to explain observed differences and to identify operational conditions that enable sustained improvement. [Results/Conclusions] Results show that DeepSeek performs well in routine, rule-based consultations. It substantially improves response timeliness and expands service availability, with an overall satisfaction rate of 86.7%. Deep integration with local library knowledge bases is associated with a marked increase in accuracy for library-specific questions (from 64.3% to 93.7%), improved precision in professional literature recommendations (by 35.2%), and higher efficiency in handling complex academic consultations (by 43.8%). However, effectiveness varies systematically: outcomes are better in science and engineering domains and in factual inquiries than in humanities and social sciences and in research- or innovation-oriented inquiries that require domain judgment and verifiable evidence chains. Mechanism analysis indicates that reliability gains depend on 1) robust local knowledge governance with version control and evidence-first retrieval, 2) scenario-specific templates and graded escalation procedures that standardize outputs by task type, 3) human-machine collaboration that supports librarian review, structured correction, and "write-back" updates, and 4) feedback-driven iteration supported by monitoring metrics and accountable operations. The study also acknowledges limitations: the sample is regionally bounded and may overrepresent early adopters; several measures rely on self-reports and short observation windows; and causal identification is constrained by cross-sectional design and rapid model/version iteration. Future research should expand to multi-region samples, incorporate longer-term operational logs, and employ quasi-experimental designs to strengthen causal inference while addressing privacy, compliance auditing, and sustainable governance in library AI services.

[Purpose/Significance] Disruptive technologies are a core force reshaping the industrial landscape, but their inherent market uncertainty contradicts traditional management logic, posing a significant challenge to the resource allocation decisions of innovation entities. To address this issue, this study, starting from the market characteristics of disruptive technologies, utilizes a hierarchical analysis framework and combines deep learning methods to identify multidimensional market demand themes for potential disruptive technologies and conduct evolutionary analysis. This aims to provide a reliable basis for strategic decision-making and resource allocation by various innovation entities, moving from "experience and intuition" to "scientific foresight." [Method/Process] Based on the substitutive market characteristics of disruptive technologies, a hierarchical analysis framework of "substitutability assessment - multi-entity demand mining - deep clustering" was constructed to identify and analyze multi-dimensional demand market themes based on potential disruptive technologies. First, the set of potential disruptive technologies that has been widely defined in existing research was systematically reviewed. Based on this, an innovation diffusion model was used to quantitatively assess their market substitutability, thereby identifying disruptive technologies with market substitution potential. Secondly, based on the identified technologies with market substitutability, and considering the demand-driven, technology transfer, and institutional guarantee mechanisms for disruptive technology market applications, this study explores multi-dimensional demand content from multiple perspectives, including users, enterprises, and government. It integrates various deep learning methods, such as user demand analysis based on multi-dimensional feature fusion, enterprise demand analysis based on text similarity networks, and government demand analysis based on data augmentation, to differentiate and mine multi-dimensional demand content. Finally, based on the mined multi-dimensional demand content, deep clustering was used to identify core market demand themes for disruptive technologies from multi-source data from users, enterprises, and government, and to analyze their dynamic evolution patterns. [Results/Conclusions] Taking the field of artificial intelligence as an example, this empirical study identified 30 potential disruptive technology market demand themes for 2021-2025, covering global digital trade technology, online behavior governance technology, intelligent waste sorting technology, intelligent transportation technology, intelligent voice interaction technology, digital cultural tourism technology, green technology innovation, green city construction technology, and smart logistics technology. The identified results have been verified by global policy documents and expert authorities, and are highly consistent with the development trends of potential disruptive technologies, effectively echoing the core directions of the current national science and technology innovation strategy and industrial transformation and upgrading. However, this study only focuses on the field of artificial intelligence and does not comprehensively cover different technological fields. Future work will extend to other technological fields to test and improve the general theory of identifying disruptive technology market themes.

[Purpose/Significance] With the growing demand for public health literacy and the accelerating digital transformation of libraries, medical science popularization services in libraries are expected not only to disseminate information but also to promote learning, skill acquisition, and behavioral change. However, existing services still rely heavily on text-based explanation, one-way communication, and short-term activity organization, which makes it difficult for users to translate received information into practical health knowledge and sustained health action. From this perspective, the core problem is not simply a lack of content or channels, but the weak connection between medical knowledge, bodily experience, everyday situations, and long-term social support. Drawing on embodied cognition theory, this study introduces the metaverse as a possible service environment for reconstructing library-based medical science popularization. The main innovation of this study lies in shifting the analytical focus from media form to cognitive mechanism, and in proposing a framework that connects scenario construction, multisensory interaction, collaborative participation, and service evaluation. This framework is expected to enrich the theoretical discussion of library health science communication and provide an operable path for the upgrading of medical-themed science popularization services. [Method/Process] This study adopts a qualitative and conceptual research design that combines literature review, theoretical analysis, and case-based interpretation. First, previous studies on library health information services, medical science communication, user adaptation, digital health literacy, and immersive technologies were reviewed in order to identify the major problems of current medical science popularization services in libraries. Second, embodied cognition was used as the core theoretical lens to extract three key dimensions, namely hybrid physical-virtual space, multisensory interaction, and collaborative community network. Based on these dimensions, the study constructs a metaverse-based service framework and explains how medical knowledge can shift from abstract presentation to contextualized understanding, embodied rehearsal, and behavioral reinforcement. Third, an immersive interactive exhibition on myopia prevention was selected as an illustrative case. The case is not used as strict empirical verification, but as a representative scenario through which the proposed framework can be mapped onto concrete design elements, including space organization, positional interaction, dynamic rendering, experience guidance, and the possibility of extension toward routine library services. This method is appropriate because the research topic is still in an exploratory stage, real-world library cases remain scattered, and conceptual clarification is necessary before controlled empirical testing and large-scale implementation can be meaningfully developed. [Results/Conclusions] The study identified three closely related bottlenecks that current library-based medical science popularization services face. First, knowledge is often detached from real-life situations. This means that users may understand medical terms superficially but still fail to apply them to concrete health decisions. Second, interaction is often limited to reading, listening, or watching, while repeated practice, correction, and embodied rehearsal are insufficient, making it difficult to internalize operational knowledge. Third, many existing services remain event-oriented and discontinuous, lacking stable support structures that connect librarians, medical professionals, users, families, schools, and communities. In response, this study proposes a metaverse construction scheme centered on three modules. The first is hybrid physical-virtual space, which organizes high-frequency health issues into explorable scenarios and links physical library space with digital simulation environments. The second is a multisensory interaction system that transforms medical concepts into visible, audible, touch-responsive, and action-related experiences, thereby strengthening comprehension through perception-action coupling. The third is a collaborative community network that extends science popularization beyond one-time events by incorporating expert consultation, peer support, family co-learning, and community participation. These three modules are integrated through a closed-loop operational logic of immersion, interaction, feedback, and adjustment. On this basis, the study further proposes implementation strategies concerning user segmentation, multimodal resource integration, platform construction, and a multidimensional evaluation mechanism covering participation, knowledge acquisition, behavioral conversion, and experience-based trust.

[Purpose/Significance] The rise of Generative Artificial Intelligence (AIGC) has made "prompt literacy" a crucial skill for effective human-AI interaction. However, there are significant gaps in public competency that risk widening the digital divides Libraries, as foundational institutions for literacy and access, are ideally positioned to address this need. This study aims to clearly define the core roles of libraries in cultivating public prompt literacy and to develop a practical, actionable framework to guide their efforts in the AIGC era, thereby enhancing their social relevance and service impact. [Method/Process] This research employs a qualitative, multi-stage approach. First, a comprehensive literature review was conducted to analyze and synthesize the theoretical conception and multi-dimensional structure of prompt literacy. Second, through a strategic analysis of libraries' inherent functions and societal mandates, the study systematically proposes a tripartite role orientation. Third, building on this role definition, an integrated practical framework was constructed. This framework synthesizes insights from library science, educational design, and technology ethics, and is informed by an examination of early innovative practices from libraries globally, moving from conceptual roles to actionable strategies. [Results/Conclusions] The study concludes that to effectively foster public prompt literacy, libraries must consciously adopt and integrate three core roles. First, as an educational guide, libraries must transition from information providers to facilitators of critical thinking and technical skill-building, specifically in human-AI collaboration. Second, as technology adapters, they must act as crucial intermediaries, assessing, curating, and sometimes tailoring AI tools to lower access barriers and meet diverse user needs. Third, as an ethical guardian, they have a responsibility to navigate the risks associated with AIGC, such as misinformation, bias, and privacy concerns, thereby fostering a trustworthy information environment. From this integrated role orientation, a detailed four-dimensional practical path is formulated. 1) Resource construction involves building a multi-layered support system, including a repository of reusable prompt templates for common and discipline-specific tasks, as well as educational materials highlighting ethical pitfalls and case studies. 2) A hierarchical education system requires the design and delivery of differentiated instructional programs. These programs range from gamified workshops for youth and students, to advanced, discipline-integrated training for researchers and professionals, and from patient, needs-based, low-barrier tutorials for seniors to programs for the digitally disadvantaged. 3) Service integration emphasizes the importance of seamlessly embedding prompt literacy support into core library services and user workflows. This includes integrating prompt design assistance into research consultations, embedding literacy modules into academic course curricula in partnership with faculty, and demonstrating AIGC applications in everyday life through community programs. 4) Ethical regulation requires the operationalization of ethical principles through explicit policies for library AI use, transparent communication with users about AI-assisted services, the development of ethical checklists and assessment tools, and the fostering of community dialogue on AI ethics. This comprehensive framework gives libraries a strategic roadmap for translating the importance of early prompt literacy development into practical, long-lasting, services. Implementing this approach allows libraries to strengthen their public education mission in the digital age, establish themselves as vital and adaptable community hubs, and play a pivotal role in fostering a more literate, equitable, and ethically conscious society amid rapid AI advancements. Future research could focus on assessing the impact of these interventions and identifying the skills necessary for librarians to fulfill these new roles successfully.