［Purpose/significance］ Based on the research on open source intelligence (OSINT) at home and abroad,this paper summarizes the theoretical and application achievements and shortcomings in order to provides guideline for future research［Method/process］ After visualizing the annual publishing trend and analyzing keyword cooccurrence of literature with quantitative method,we further concludes the concept of OSINT and summarizes the research content and structure combined with the DIWK model［Result/conclusion］ OSINT has been extensively explored,especially in the concept,collection,fusion,quality evaluation,model construction,etcHowever,the research on OSINT is still in its infancy,the research on theoretical models and methods is relatively lessIn addition,the application scenarios and content still need to be expanded

［Purpose/significance］ Persona is a fast and precise data analysis tool to analyze user behavior patterns,consumption habits and other business information,which can lay the foundation for precise marketing and user experience improvement ［Method/process］ By the method of literature analysis,from the dimensions of definition,constituent elements,models and methods,this paper carries out a careful review of the latest research results of persona,points out the different modeling methods of persona,sums up the characteristics of these methods and the future direction of development,finally shows the implications of persona on Library and Information Science ［Result/conclusion］ As an emerging data analysis tool,its theoretical study and practical application are still relatively weak,so there is still a lot of room for exploration Through the literature review,the paper finds out the direction for further exploration and provides references for further research

［Purpose/significance］ Technology transaction is the key link of technology transferIn addition to the matching of technology content,technology transaction is affected by many factorsStudying the influence mechanism of supply and demand matching in technology transaction is conducive to technology transaction［Method/process］ Based on the data of Hubei Technology Exchange’s past trading contracts,this paper combs five antecedent variables of technology,subject and environment based on the TOE framework,and applies QCA method to explore the condition configuration and path of high turnover rate technology trading［Result/conclusion］ It is found that there are two driving mechanisms for high turnover rate of technology trading,namely,the subjective type of distance policy perceived reliability and the objective type of distance policy organizational proximity technological reliabilityIn the two paths,distance and policy are important factors,while subjective reliability and objective technical conditions can produce substitution effect to a certain extent

In recent years,under the trend of “traffic economy”,social media platforms have excessively used algorithmic technology to push users into “information cocoons” and “echo chambers”The extreme views caused by group polarization have increased significantlySo the study in the formation mechanism of group polarization on the platform can provide a reference for the government to regulate the polarization of public opinionOn the basis of reviewing relevant domestic and foreign research literature on group polarization,through the observation of the case,the study summarizes the social effects of network polarization and the causes of group polarization in social media platformsBased on previous studies and summarized factors,the study analyzed the three combined forces of network group polarization,and constructed the dynamic mechanism model and formation mechanism of the formation mechanism and guiding strategies of group polarization in Social Media PlatformsBased on the obtained model and the conclusions,the study proposes corresponding guidance strategies.

Abstract:The construction of emergency intelligence system is a powerful basis for management and decisionmaking after major public health events.However,there are still some deficiencies in the current emergency intelligence system in the aspects of intelligence storage,intelligence sharing,intelligence traceability,rapid warning and linkage response.Therefore,based on the current emerging blockchain technology,this paper introduces the ideas of blockchain,such as nontampering,decentralization,traceability and selftriggering,into the construction of emergency intelligence system.We use blockchain to store,share and trace intelligence integrally,uses smart contract to set early warning threshold,and broadcast all over the network to generate gear linkage response mode.According to the different stages of the development of the epidemic,we improve the emergency information system under the current major public health events,so as to provide the basis and guarantee for improving the national emergency management ability.

［Purpose/significance］ User portrait describes the behavioral characteristics of the user group of a specific product，and provides merchants with multifaceted user information，so that merchants can understand their customers，improve products or services，and bring possibilities for precision marketing［Method/process］ Based on the online review，the conceptual model of user portrait is constructed from three dimensions：user information attribute，hotel information attribute and user evaluation information attributeThen the tool of Protégé is used to establish the ontology to realize the association between user portrait attributes，and accomplish the complete description of the hotel users’ characteristics［Result/conclusion］ Taking the online reviews of Ctrip hotels in Beijing as an example，the dimensions of user，hotel，and comment information are visualized to build user portraitThe result provides basis for merchants to understand customers’ needs and implement the strategy of precision marketing

［Purpose/significance］ On the basis of domestic research,this paper systematically sorts out related literatures on data fusion and integration abroad,aiming at providing references for the work in the field of data fusion and integration［Method/processes］ This paper mainly reviews relevant research,research approaches and methods,and applications and challenges,and points out future research directions from three aspects:concept,path and application［Result/conclusions］ Research approaches and methods include data fusion and integration model and framework,phasebased strategy,featurebased representation method and semanticbased applicationAnd it is found that there are common or special application challenges in each field

［Purpose/significance］ The individuals are becoming “naked people” in the era of social media,which reflects that exploring the determinants and consequences of privacy attitude is necessary［Method/process］ Based on social cognitive theory and “cognitiveattitudebehavior” model,this study integrates both of internal factors and external elements and utilizes the Structural Equation Model (SEM) and FuzzySet Qualitative Comparative Analysis (fsQCA) to investigate the mechanism of digital natives’ intention to protect privacy［Result/conclusion］ The SEM results reveal that privacy attitude could be affected by privacy literacy and information sensitivity;peer effects only could positively affect privacy concern;parental influence could not exert any impact on privacy attitude;also,the privacy fatigue plays triple roles in the formation from privacy concern to intention to protect privacyThe fsQCA results indicate that four and three configurations lead to the high/low level of intention to protect privacy［Limitations］ Only quantitative method was used in this research,the future studies can apply the qualitative method to explore the deep reasons of such phenomenon

［Purpose/significance］ The impact evaluation of academic papers is an important methodological issue in the evaluation of academic achievements,and it is also an important practical topic that needs to be improved in scientific and technological evaluation or management activitiesThe use of ecological niche theory can provide a new research perspective and method for the impact evaluation academic papers from the perspective of resource occupation,so as to enrich and develop the evaluation theory of academic achievements［Method/process］ According to the niche theory,this article divides the influence of academic papers into primary influence and regenerative influence,based on the perspective of resource occupation of the entire academic exchange ecosystem,and constructs an evaluation model for the niche breadth of academic papers from the three dimensions of content innovation,online attention and offline attention,so as to assess the impact of academic papersAnd select academic papers in library and information science for empirical analysis,to discusses the impact evaluation method of academic papers based on the niche perspective［Result/conclusion］ The experimental results in this article show that the academic paper influence evaluation model proposed in this paper based on the niche theory,can be used as an effective method to evaluate the influence of academic papers scientifically and reasonably

［Purpose/significance］ The research deepens the perspective of knowledge graph construction into the innovative content contained in patent documents,so as to focus and express the core innovation information of each patent structurally in a specific field,briefly express how the patent individual achieves innovation,reveal the essence of each patent innovation and semantic network of patent innovation in the field［Method/process］ The research uses the Chinese patent documents abstract in the field of educational robotics as a dataset to build a knowledge graph of patent documents innovative contentBert model is used to identify and extract innovative sentences in patent documentsInnovation topic was analyzed by LDA topic modelBased on Bert4keras,entities and their relationships are extractedThis research uses Neo4j to store the knowledge graph［Result/conclusion］ The deep learning method adopted in this study can effectively discover the core content of patent documents innovation,and the patent innovation knowledge graph constructed can better summarize and organize the core information of patent,which is technological innovation information

［Purpose/significance］ This paper aims to analyze the characteristic performance,the antecedents and outcomes of information cocoons for theorization of information cocoons.［Method/process］ The existence of information cocoons is still controversial and the concept is not yet clear.For this reason,this paper conducts an in-depth analysis of information cocoons,collects research data using in-depth interviews,analyzes interview data using grounded theory,and builds a theoretical model of information cocoons based on the coding results.［Result/conclusion］ The paper verifies the existence of information cocoons and extracts the main characteristics of information cocoons;summarizes the influencing factors of information cocoons,such as personalized recommendation,individual characteristics,environmental characteristics;discusses the influence of information cocoons in the individual and group sense respectively.［Limitations］ The theoretical findings from qualitative analysis should be tested by subsequent empirical studies,such as using structural equation model to verify the influencing factors of information cocoons.

［Purpose/significance］ Humanities scholars face the technical dilemma of using traditional means to perceive data and the humanistic illusion of interactive cognition of data in the era of big dataFor spreading Chinese traditional culture and telling Chinese stories well,regardless of producerdriven or consumerdriven modes,it’s necessarily to pursue an information exchange and communication tool with both technical and humanistic frameworkData Storytelling (data narrative) can be endowed with new usage context and meaning in this sense,and supported by data science theory and practice,so as to become an important node of digital humanities methodology even technology system［Method/process］ The research starts from the data perception and data interaction cognitive dilemmas in the era of big data,and clarifies the concept and logical context of data narrative through literature surveys,and refines the basic elements and theoretical foundations of data narrativeThen compares the application scenarios of early fields of data storytelling,such as media,business,education,etc,and in which analyzes the value of user experience,emotion,knowledge perception and communication,and cognitiondriven value of data storytelling,and then summarizes the application and practical significance of data narration from the perspective of Digital Humanities;finally puts forward the path and method of digital humanities for research and education by using data storytelling［Result/conclusion］The exploration and development of digital humanities in different languages and cultures have been iteratively upgraded around the technical law of data awareness and value discovery,as well as the humanistic framework or paradigm of narrative discourse systemUsing data science as a platform,information visualization as product or service form,and narratology as channel for value realization,data storytelling will be beneficial to promote the development of digital humanities,and become an important tool for integrating data resource,even spreading Chinese culture and stories

［Purpose/significance］ Identifying and exploring the innovation content of different granularity in the full text of scientific and technical papers can reveal the “past life” and “history” of theories,methods and technologies in a certain innovation content,which can help researchers in the process of scientific and technical innovation to The article is based on deep learning model,semantic model,and theoretical model,which can help researchers to obtain the required scientific and technological innovation intelligence,quickly grasp the innovation dynamics in the field,select the innovation direction,determine the innovation plan,and improve the efficiency of scientific and technical innovation.The article explores the innovation segment localization method based on deep learning model and semantic role information,and identifies important functional sentences,so as to lay the foundation for the construction of the innovation knowledge map of scientific and technical papers.［Method/process］ The full text of scientific and technical papers from intelligence journals is used as the research object,and the innovation sentences in the full text are identified by BERT pretraining model.The innovation sentences are used as the basis to match the innovation paragraphs by semantic role information.The important innovationrelated function sentences in the innovation paragraphs are identified by guide words.［Result/conclusion］ The experimental results show that the semantic informationbased innovation paragraph identification can effectively expand and improve the content of innovation sentences,and the functional sentences present in the innovation paragraphs are mainly the discussion of “innovation process” and “innovation contribution”.The experiment proves that the method proposed in the paper is feasible and effective in mining and extracting the innovation content of scientific and technical papers with different granularity.

［Purpose/significance］ Combining coreperiphery theory and patent citation analysis,a new method to identify potential disruptive technologies was proposed to help reveal and judge the technical directions or topics that may have disruptive effects,so as to optimize R&D layout for government and enterprises and win competitive advantage［Method/process］ Taking American authorized patents in the field of artificial intelligence between 2000 and 2019 as samples,based on the patent and scientific references in patent documentation,the technologyweighted and scientific knowledgeweighted cooccurrence network were constructed respectively,and the citation relationship between the two networks was formedThe coreperiphery theory was used to design the candidate criteria for disruptive potential technology nodesThe time series analysis method was used to determine the disruptive technology from three dimensions,which were technology foundation,influence and mutation［Result/conclusion］ The results show that combining the coreperiphery theory and patent citation analysis can effectively identify disruptive technologies in specific technical fieldsSome fields in artificial intelligence,such as machine learning,image recognition and processing,AI medical care,and autonomous driving technologies,have disruptive potentialityAnd these technologies are closely related,highly permeable,and have an obvious trend of coordinated development

［Purpose/significance］ This paper aims to expound largescale knowledge graph and its typical application,analyze the challenges and countermeasures it faces,and present its development trend ［Method/process］ Using the methods of network investigation and content analysis,the paper systematically explores the largescale knowledge graph and its typical applications,its challenges and countermeasures,as well as its development trends ［Result/conclusion］ The typical applications of largescale knowledge graph are semantic search,indepth question and answer,intelligent recommendation,etc It faces the challenges of feature knowledge representation,knowledge lacking,knowledge updating,function comprehensiveness,and so on It is developing in the direction of specialization,openness and intelligence

［Purpose/significance］ Practice of management requires a deep understanding of user needsIn addition，as a significant channel of user preference expression，online review is promising source from which to reveal user needsTherefore，how to identify user needs from massive online reviews become an important issue［Method/process］ In this case，based on KANO model，we propose an analytical framework of user needs for online reviewsSpecifically，product attributes are first elicited from online reviews based on LDAAfterwards，the short sentences are labeled into different attributes using BERT model，and the sentiment orientations of the short sentences are identified at the same timeNaturally，user needs can be identified by KANO model according to the short sentences and sentiment analysisBesides，a cell phone review dataset is used to validate the proposed model［Result/conclusion］ The framework divides product attributes into four categories:mustbe quality attribute，onedimensional quality attribute，attractive quality attribute，and indifferent quality attribute，which will be presented in visualizationThe empirical study show that user needs can be effectively identified，and improvement strategies can be provided for enterprises［Limitations］ The dataset in this paper is limited and unbalanced

［Purpose/significance］ Digital twin technology brings new opportunities for digital economic growth and smart cities constructionAs an important part of smart city,the smart library scenario services based on digital twin technology will be a main direction for future transformation and upgrading［Method/process］ By analyzing the concept and characteristics of digital twin technology,we propose the application scenario and architecture of digital twin technology in smart libraries,and combine with the“scenarioinformation expectationsituation”,explore the path and model of scenariobased services for smart libraries with digital twin technology［Result/conclusion］ Digital twin technology has a wide range of application scenarios in smart libraries,and can provide better scenariobased services for users through the bidirectional iterative optimization of virtual libraries and physical librariesStandardization of construction,sharing of information,professionalization of developing talents and rationalization of privacy will be the development trend of innovative practice of scenariobased service of smart libraries with digital twin technology

［Purpose/significance］ The current research literature on topic evolution and prediction growth rapidlyTimely review of relevant research will help enhance the research value［Method/process］ Firstly,we summarize the research theory foundation and draw the research frameworkAnd then,we analysis the methods of topic evolution and topic prediction separately.Finally,we discuss the value of the research,point out the problems and put forward suggestions for future researchers［Result/conclusion］ Subsequent research needs to strengthen the application of scientific evolution model,evolutionary diffusion theory and sciencetechnology linkage model,actively use language model and deep learning algorithm to improve the accuracy of topic prediction,enrich the research at macro,meso and micro levels,and pay attention to the effectiveness of scientific and technological intelligence

［Purpose/significance］ As for the complex and changeable characteristics of national security scenarios,multisource information needs to be fused for situational awareness and situation analysis of national security events,which could provide comprehensive,realtime and accurate data support for emergency management of national security events［Method/process］ We first analyze the intelligence sources of national security events from the perspective of the threeworld theory;then,we propose an eventcentric multisource information collaborative awareness mechanism as well as a situation analysis framework based on the “timespacetopic” model;finally,we developed a sample situational awareness and analysis system for national security events based on the methodology proposed in our paper［Result/conclusion］ With the help of information organization and information analysis theories and methods,the orderly organization and multidimensional analysis of situation information related to national security events can promote the effective use of situation information,thereby improving the efficiency of emergency response to national security events

［Purpose/significance］Exploring the formation mechanism of discontinued behavior of strongties social media users and the interaction mode of variables so as to provide comments and suggestions for strongties social media developers and operators to improve service quality,optimize user experience and improve user stickiness［Method/process］With the typical application of strongties social media WeChat Moment as the research object,this article gets the text data by depth interview and adopts grounded theory to extract concepts and categories through open coding,axial coding and selective coding,and builds the conceptual model under the framework of SOR theory［Result/conclusion］ The situational factors (internal factors of the platform) and environmental factors (external factors of the platform) stimulate users to generate negative emotional cognition,which leads to different levels of discontinued behavior,and personal factors consisting of demographic factors,personality traits,usage characteristics,and relationship characteristics modulate the discontinued behavior of strongties social medias users

［Purpose/significance］ With the widespread application of artificial intelligence in the economy and society,it has become a global consensus to strengthen the regulation and governance of artificial intelligence and control its inherent risksIt is of great significance to investigate the regulatory design of major countries and regions in the world for improving our artificial intelligence regulatory system［Method/process］ Through literature analysis and comparative analysis,the concept of artificial intelligence regulation and the risks of artificial intelligence application are defined,and on this basis,the design of artificial intelligence regulation in major countries and regions in the world is introduced,and the development trend of artificial intelligence regulation is summarized［Result/conclusion］ The global competition for artificial intelligence regulation has begun and the degree of rule of law is deepeningArtificial intelligence regulation based on technical risk classification has become a mainstream trendChina should also strengthen the toplevel design of artificial intelligence regulation,adopt the regulatory thinking of classification,actively participate in international cooperation in artificial intelligence regulation

［Purpose/significance］ In order to clarify the causative factors of information cocoon towards researchers under the Internet environment,prevent the emergence of academic information cocoon,and create a good information environment for researchers［Method/process］ Used the grounded theory methods to make qualitative analysis by conducting semistructured interviews,data coding and model development［Result/conclusion］ The results show that internal factor of researches,information factors,social environment and information situation factors are the key factors affecting the formation of information cocoon towards researchersSome suggestions,such as improving scientific management and services,enhancing interdisciplinary academic exchange and cooperation,and promoting information literacy,were pointed out in the end

［Purpose/significance］ To study the influencing factors and configuration path of digital hoarding behavior with college students as a sample,which can help enrich the related research in this field and provide some suggestions and countermeasures for the intervention of digital hoarding behavior among college students.［Method/process］ Based on social cognitive theory,we analyzed the influencing factors of digital hoarding behavior from both individual and situational perspectives through questionnaire survey method and fuzzy set qualitative comparative analysis method and constructed a theoretical model.［Result/conclusion］ The necessity and sufficiency analysis was conducted by fsQCA software and R software,and the results concluded that there were no necessary conditions for digital hoarding behavior among the six condition variables,and a total of five paths were obtained to trigger digital hoarding behavior,which can be further integrated into three major configurations according to their characteristics and the distribution of core conditions:preventive measures,precise location,and element aggregation.

［Purpose/significance］ Knowledge community serves as a platform for highdensity sharing and innovation of scientific and technological knowledgeHow to effectively engage users in topics about scientific and technological innovation is jointly concerned by knowledge communities and public cultural institutions［Method/process］ Based on the theoretical framework of MOA,questionnaires were given out among knowledge community users and relevant survey data was collectedFuzzyset QCA was used to explore the configuration effects of motivation,opportunities and ability on knowledge community users’ participation in scientific and technological innovation topicsA staged participation behavior model of knowledge community users in scientific and technological innovation topics was established based on the core configuration results［Result/conclusion］ There are three types of configuration modes influencing knowledge community users’ participation behavior in scientific and technological innovation topics,namely the high community support and individual efficacydriven mode,the high network connection coredriven mode and the high innovationdriven modeMotivation,ability and opportunities affect the acquisition stage,reception stage and participation stage of topics,respectivelyUsers’ participation in scientific and technological innovation topics is significantly affected by the degree of network connection

［Purpose/significance］ To explore the basic characteristics and internal regularity of the spread of network false information over time under the infodemic,to provide some theoretical support for the management of the spread of network false information under the infodemic,and to provide an effective way for relevant departments and platforms to curb the spread of network false information［Method/process］ Using the crawler program in python language to collect the data set of network false information about the novel coronavirus pneumonia,constructing the time series of network false information dissemination,using statistical analysis,text classification,and topic mining technology,from the quantity,source and subject dimension to explore the dynamic transmission characteristics and regularity of network false information under the infodemic over time［Result/conclusion］ The study shows that the network false information under the infodemic presents a segmented evolution,videoization and repeated fluctuations in the spreading regularity,and the COVID19 infodemic paralleled the propagation of the COVID19 epidemicIts sources and topics are different in the different stages of time series