[Purpose/Significance] In order to improve the efficiency of emergency information flow, this study proposes an agile emergency information system. [Method/Process] This paper analyzed the realistic predicament of the agile emergency information system from three aspects: the physical world, the information space and the social system, based on the theory of agile governance, and it analyzed the connotation of agile emergency inf-ormation system, and put forward an agile emergency information system architecture. [Result/Conclusion] The agile emergency intelligence system mainly faces the dilemma of aggregation, integration and immediacy.Digital intelligence empowerment promotes the agility of the emergency intelligence system by formulating a whole-process emergency data absorption mechanism, establishing a linked emergency intelligence community, and using diversified emergency intelligence technologies.The agile emergency intelligence system has three core concepts: sensitivity and speed, connectivity and consensus, and two-way feedback, and realizes the operation mechanism of real-time risk perception, rapid information generation and scientific information provision based on the three interactive systems, so as to improve the efficiency of emergency management.

[Purpose/Significance] Under the profound integration of digital and intelligent technologies with information systems, as exemplified by the"cloud computing services-the Internet of Things-blockchain-big data-artificial intelligence-mobile networks"paradigm, intelligence digitalization has transformed the traditional modes of information flow and interaction.This transformation has resulted in a paradigm shift within the information ecosystem, ushering in a more intelligent, interconnected, secure, and sustainable digital and intelligent (DI)information ecosystem.Thus, a deep understanding and grasp of the DI information ecosystem are of paramount importance for the advancement of information digitalization and intelligence. [Method/Process] Based on the theory of information ecology, this study decoded the fundamental concepts of the DI information ecosystem.It elucidated its connotations and conceptual framework, analyzed its constituent elements, and dissected its operational mechanisms, provided a systematic interpretation of the DI information ecosystem. [Results/Conclusion] The DI information ecosystem is a contemporary manifestation of the evolution of information ecosystems.It is composed of four key elements: digital and intelligent information, platforms, entities, and environment.Integrating the theory of information ecology, this study dissects the operational mechanisms supporting the effective functioning of these four elements.

[Purpose/Significance] It is of great significance for rational selection of topic modeling technology to carry out topic mining in scientific literature to study and compare the application performance of different topic modeling methods and tools in the topic recognition of scientific literature. [Method/Process] By constructing experimental corpus of chinese and English scientific literature, three topic modeling methods(LDA, Top2vec, Bertopic)and five document feature representation methods(Bag of Words, TFIDF, Doc2vec, MiniLM, SciBert)were selected for experiments.And the index of topic diversity, topic coherence, topic stability and topic variability of different modeling methods were compared and analyzed. [Result/Conclusion] Among the topics identified by different modeling tools, LDA and Bertopic accounted for 9.81% and 7.46% of topics with similarity relations in English and Chinese corpora; Top2vec model based on Doc2vec algorithm has the best performance on topic diversity index, and top2vec model and bertopic model based on text pre-training algorithm have better topic stability and dispersion index than traditional topic modeling methods.In view of the rapid development and wide application of the current large language model technology, accelerate the development of scientific literature pre-training models and apply them to scientific and technological information Business practice is an important research direction at present.

[Purpose/Significance] Knowledge diffusion and dissemination can promote knowledge inheritance and innovation.Exploring the characteristics of scholars' knowledge diffusion at the content level is helpful to reveal the role of scholars in the development of subject knowledge, which improves the evaluation of scholars' academic contributions. [Method/Process] This paper collected 105 academic papers and citations of Professor Liu Zeyuan as the first author, and explored the characteristics of Professor Liu Zeyuan's knowledge diffusion through citation content analysis from the three dimensions of discipline field, knowledge unit and theme thought. [Result/Conclusion] The research results show that in the discipline dimension, the cited literature covers 82 disciplines, which shows the wide influence of Professor Liu Zeyuan.In the knowledge unit dimension, 98 non-repetitive knowledge units are extracted from the citation sentences, including concept, theory, and method and technology.In the theme thought dimension, the five thematic knowledge studied by Professor Liu Zeyuan has been inherited and developed in different disciplines.This multi-dimensional study of scholars' knowledge diffusion helps to reveal the role of scholars in the development of knowledge of different disciplines in an all-round way, and provides a direction for evaluating the influence of scholars' knowledge.

[Purpose/Significance] As a passive information behavior, information avoidance will affect the sustainable development of social media platforms.Based on the cognitive dissonance and psychological empowerment theory, this research integrated a dual perspective of enablers and inhibitors to explore the antecedents and development paths of social media users' information avoidance behavior. [Method/Process] A mixed method of structural equation model and fuzzy-set qualitative comparative analysis method were adopted for analysis. [Result/Conclusion] The results show that social support and information quality have a significant impact on psychological empowerment, which in turn negatively affects information avoidance behavior.Social comparison and perceived overload affect the cognitive dissonance, which further leads to information avoidance behavior.The fsQCA results find three paths that trigger users' information avoidance behavior.The results suggest that social media platforms should adopt effective information filtering mechanism and create supportive platform climate, thus mitigating users' psychological dissonance and reducing their information avoidance.

[Purpose/Significance] This paper aims to explore the antecedents and consequences of graduate students' academic digital hoarding behavior. [Method/Process] Taking 20 graduate students as subjects, the study conducted the semi-structured interviews so as to obtain the research data, and then analyzed the data by utilizing the grounded theory method. Through open coding, axial coding and selective coding, 44 initial concepts, 16 categories and 8 main categories were summarized, thereby the antecedents and consequences of graduate students' academic digital hoarding behavior were discovered. [Result/Conclusion] Results show that the antecedents of graduate students' academic digital hoarding behavior include personal-related factors(individual differences and motivations), academic digital resource related factors(qualitative, quantitative, and acquisition characteristics), and environmental/situational factors(technological, social, and task situation factors). Meanwhile, graduate students' academic digital hoarding behavior could result in psychological consequences(cognitive and emotional effects), behavioral consequences(impact on personal academic information management behavior and scientific research task completion), economic consequences(expenses for storage device purchase and costs of online storage service), and social consequences(establishing a good social image and improving interpersonal relationships).

[Purpose/Significance] To extract user demand of clinical decision-making driven by electronic medical record data, and construct a demand model, under the perspective of information chain. To bridge the gap between clinical decision support services and actual clinical work needs, to provide value release targets for electronic medical record data, to expand the application domain of information chain, to help the construction of data-driven clinical decision support platform, and to provide guidance for clinical intelligence services. [Method/Process] Using the method of template analysis, through the analysis of interview, 7 level-1 need themes, 24 level-2 need themes, 54 level-3 need themes, 43 level-4 need themes and 2 level-5 need themes were refined, and a hierarchical model of clinical decision-making driven by electronic medical record data was constructed, based on the theory of information chain. [Result/Conclusion] The demand of users for clinical decision-making driven by electronic medical record data can be summarized along the chain of information, including(1)intelligent recording of electronic medical records, (2)organization and extract of clinical key information, (3)disease risk prediction driven by electronic medical record, (4)supplement and discovery of disease knowledge, (5)automated assistance in diagnosis driven by electronic medical record, (6)cause analysis and prompts of abnormal conditions, and(7)assist in the development and recommendation of treatment plans. Doctors showed different levels of enthusiasm for the application of electronic medical record data-driven clinical decision making, and expressed their concerns about the risks, such as unclear data rights and low information technology maturity.

[Purpose/Significance] The paper aims to discovery the characteristics of data reuse behavior of scholars in education, and helps libraries and other institutions develop more targeted data service strategies in the process of scientific data management and service. [Method/Process] The paper used content analysis to analyze the scientific data reuse behavior in 544 fund-sponsored papers from 2017 to 2018. [Results/Conclusions] The paper finds that data reuse ratio of scholars in education is relatively high and stable, and it is more inclined to reuse survey data. In terms of data reuse sources, government websites, data centers and journal papers are the main sources of data reuse for scholars in education, and there are some problems such as less secondary analysis of scientific data and data reuse specifications are inadequate.

[Purpose/Significance] The primary objective of this study is to comprehensively understand user requirements for a medical and health application tailored to address the challenges associated with aging. [Method/Process] Through a systematic review of literature and expert investigations, 26 demand attributes were identified and categorized using the Kano model, mixed-class analysis, and Better-Worse coefficient analysis. The categorization revealed eight essential attributes, seven expectations, four charm attributes, and seven attributes without significant distinction. [Result/Conclusion] Recommendations were formulated to enhance essential supply, improve expectation quality, expand attractive supply, and adjust content perceived as having no significant difference. The overarching goal is to effectively align with user demands through a hierarchical approach, focusing on optimizing health applications for aging individuals and facilitating their seamless integration into modern society. The insights generated from this study provide valuable reference points for the ongoing transformation of medical and health applications, ensuring a more inclusive and user-centric design suitable for the aging demographic.

[Purpose/Significance] Self-media resources have become an important means of accessing information. Their multi-source and heterogeneous characteristics pose a challenge to effective knowledge organization.. This study aims to construct and implement a knowledge organization model for multi-source media resources. The technical paths of textualization, knowledge meta-extraction, and semantic association will be used to achieve this goal. The lack of a unified metadata standard for media resources is the motivation behind this study. [Methodology/Process] The study began by repurposing and adapting DC metadata to create a metadata description framework for self-produced media resources. It then developed a multi-origin media resource ontology and semantic network. Computer algorithms were used to standardize the text structure of multi-origin media resources, allowing for knowledge organization research to delve into the specific content of self-produced media resources. Furthermore, named entities, keywords, and knowledge summaries from self-media resources were extracted. We also conducted semantic association experiments based on data from platforms such as Bilibili, Tiktok and Zhihu. [Results/Conclusions] The algorithmic association between self-media resources and virtual collection resources enables a semantic mapping, which aims to explore the multifaceted value of self-media resources and provide new ideas for knowledge services in libraries.

[Purpose/Significance] Exploring the impact formation pattern of highly cited papers can help improve the theoretical system of evaluating the impact of a single academic paper, and it also has important reference significance on how to enhance China's academic impact and academic discourse power. [Method/Process] This study introduced the concept of "dynamic context" and explored the impact formation pattern of highly cited papers based on the dynamic process perspective of theme evolution. The dynamic context of theme evolution was constructed into three levels according to the granularity from fine to coarse. The third level performed pattern extraction, and the first and second levels assisted in pattern analysis. The gene editing field was selected as an example for pattern exploration. [Result/Conclusion] Through empirical analysis, it is found that there is one major pattern and two minor patterns in the formation of the influence of highly cited papers. The main pattern is the "inheritance-inheritance" pattern, which means that the selected topics of the papers are centered on the core issues of the field and realized progressive breakthroughs to form influence. The secondary patterns are the "hybrid-inheritance" and "hybrid-hybrid" patterns, which mean that the topic of the paper focuses on the long-term cross-cutting and fragmented thematic paths of the field, and has the potential to contribute to the breakthrough of the core issues of the field to form influence, or to form influence within the subsequent derived thematic paths.

[Purpose/Significance] Under the background of the rapid change of scientific knowledge, the irregular citation behavior has gradually evolved into a complex phenomenon of hidden and diverse forms. The identification method of irregular citations is established through the information of literature sources, and the influencing factors and internal mechanism of its formation process are revealed. It is of great significance to the standardization of scientific journals, the fairness of scientific evaluation system and the perfection of academic ecological environment. [Method/Process] Firstly, multi-dimensional evaluation indexes such as citation content similarity, use-citation conversion rate and coupling strength were constructed. Secondly, the study combined with the characteristics of relevant literature and the characteristics of citation content, explored the influencing factors behind the irregular citations. [Result/Conclusion] In this study, nearly 40 000 groups of hidden triangular citation relationships are found in 3 million biographic coupling, indicating the universal existence of hidden triangular citation behavior in the scientific community. In the citation context analysis, document language, document type and subject area are the influencing factors for literature C to quote reference A, while the influence of journals where literature A and literature B are located, their own citation influence and publication time difference are the influencing factors for literature C to choose to hide reference B. The scientific community should be aware of the urgency and importance of this kind of irregular citation behavior, and gradually establish a recognition and supervision mechanism.

[Purpose/Significance] Academic journals are one of the important carriers of academic communication.The editorial board of journals bears the important responsibility of grasping the direction of journals and checking the content and quality of papers.The research topics of editorial board can reflect their awareness and perception of academic trends in this field.This paper aims to explore the impact of research topics of editorial board on the subject of journal articles. [Method/Process] Taking the first-class journals in the field of information science as the research object, this paper constructed the editorial dataset and the non-editorial dataset, excavated the theme of the articles through the dataset, and used the Word2Vec model and the Kmeans clustering method to explore the topic distribution characteristics in the journals and the influence of the editorial board members. [Result/Conclusion] The study found that editorial board members have a higher impact on cold topics than on popular topics, and that editorial board members can guide scholars to study valuable topics by publishing papers on emerging topics.

[Purpose/Significance] Under the background of the rapid development and profound changes of artificial intelligence technology and applications, new research topics and methods are constantly emerging in the field of machine learning, and deep learning and reinforcement learning technologies continue to develop.Therefore, it is necessary to explore the evolution process of machine learning research topics in different fields and identify hot and emerging topics. [Method/Process] Taking the machine learning research papers in the Web of Science database from 2011 to 2022 in the field of library and information science as an example, this paper combined the LDA and Word2vec methods to conduct topic modeling and topic evolution analysis, and introduced topic intensity, topic influence, topic attention and topic novelty to identify hot topics and emerging hot topics. [Result/Conclusion] The results show that: ①The combination of Word2vec semantic processing ability and LDA topic evolution ability can identify the research topic more accurately, and intuitively show the stage evolution law of the research topic; ②Machine learning research topics in the field of library and information science are mainly divided into three categories: natural language processing and text analysis, data mining and analysis, and information and knowledge services.The correlation among various topics is strong and has the characteristics of theme correlation evolution; ③The theme intensity, influence, attention and comprehensive indicators designed can better identify the hot topics in three different cycle stages(2011—2014, 2015—2018 and 2019—2022).

[Purpose/Significance] PMC index model has become one of the significant methods in policy documents quantitative research.Its relevant research results are abundant, so it needs to be summarized systematically to promote the further development of this method. [Method/Process] Using bibliometric method, this paper explored the current situation of the application of PMC index model in China, sorted out the general process of PMC index model, described the practice of each step in the process, and discussed how to use the model more rationally. [Result/Conclusion] The application of PMC index model to policy evaluation is a narrow evaluation.The rational use of PMC index model needs to clarify this position, build scientific index system and ensure the consistency of policy text in application process.