[Purpose/Significance] In order to solve the problem of lack of diversification and personalization in the recommendation results of current microblog recommendation methods,[Method/Process] this paper proposed a diverse and personalized recommendation method to combine blog semantics and user interaction.Firstly,according to the semantic relationship between blogs and the interaction relationship between users who browse blogs,the paper constructed the topic-interaction correlation matrix to realize the fusion of the semantic relationship between blogs and users;secondly,on the basis of the topic-interaction correlation matrix of blogs,the k-means clustering method was used to divide blogs' diversification,and a number of different blog clusters were obtained;finally,the algorithm of matrix decomposition class was used to select and sort blogs,and the recommendation list of blogs satisfying users' preference was selected from each cluster.[Result/Conclusion] The experimental results showed that this method not only made the blog recommendation lists have diversity,but has higher recommendation accuracy,but also realized the organic integration of blog recommendation diversification and personalized recommendation.

[Purpose/Significance] We investigate the privacy behavior of different entities,establish an effective privacy protection mechanism,and promote the development of intelligent health.[Design/Methodology] Considering the current research limitations of privacy behavior,such as the uniformity of behavior interaction,without considering the bounded rationality,ignoring the dynamic of regulation behavior and how to accomplish Kaldor improvement,a novel analysis method based on evolutionary game is proposed here as follows.First,the game players consisted of intelligent health providers and regulation departments are designed,to investigate their behavior in the process of intelligent health application.Then,by analyzing the profits,costs and loss of different game behaviors,the interaction mechanism and evolutionarily stable strategies(ESS) are employed.Based on this,the impacts on evolutionary trend of different factors are discussed.[Finds/Conclusion] The results show that the bilateral privacy behavior is correlated with investment costs,marketing profits,regulation costs,reputation loss,regulation efficiency,fines and subsidies,in which regulation efficiency,fines and subsidies are the major factors affecting the interaction mechanism and evolutionary trend.Under the condition of static parameters,the interaction of bilateral privacy behaviors is linear.Under the condition of dynamic parameters,the influence of regulation behavior is weakened,and the convergence of investment behavior is mainly dependent on the reduction of costs and the increase of market profits.[Originality/Value] By applying the proposed method in this paper,the privacy regulation mechanism can be dynamically adjusted,to reduce the administrative intervention and promote the development of intelligent health.

[Purpose/Significance] In order to improve the speed of emergency response and reduce the loss of emergencies,this paper analysis the emergency scenario monitoring in the target area through the public opinion sensitive information mining,for that the online public opinion can be spread faster than the emergencies being recognized.[Method/Process] This paper extracted the feature information in the public opinion both by the statistical and classification methods.Based on that,it picked the public opinion sensitive information,and generated the mapping function between the public opinion sensitive information and the emergency scenario.Then,it mined the sensitive information of public opinion through mapping function in the target area,and conducted scene to determine the target scenario.[Result/Conclusion] Through the corresponding relationship between the sensitive information of public opinion and the scene,the emergency scene portrait can improve the effect of the emergency scene analysis in theory.

[Purpose/Significance] Exploring the influence mechanism of social media recommendation on users' online interaction intentions from the perspective of privacy is helpful to guide social media platforms in providing reasonable personalized recommendation services and ensuring the activity of users' online interaction.[Method/Process] Combined with S-O-R model,privacy computing theory and information boundary theory,the influence model of social media recommendation on users' online interaction intention was constructed from the perspective of privacy.The short video social platform was taken as the investigation scene,and the data were collected through network questionnaire,and the structural equation model was used for empirical test.[Result/Conclusion] The empirical results showed that social media recommendation had a significant positive effect on users' perceived benefit,perceived privacy risk and platform trust.Perceived benefit had a significant positive effect on perceived information control and platform trust.Perceived privacy risk had a significant negative effect on perceived information control and platform trust.Perceived information control and platform trust had significant positive effects on users' online interaction intention.

[Purpose/Significance] To analyze the main cognitive factors and their relationships involved in the process of users' health risk information seeking,to reveal the users' psychological internalization characteristics of health risk perception and the mechanism of transformation into information search behavior,in order to provide the the basis for the cognitive intellectualization of information service system.[Method/Process] Under the theoretical framework of problem-solving situation theory,the paper took COVID-19 as an example,collected data via questionnaire surveys statistical data analysis and empirical testing were used to explore the relationship of health risk perception,problem recognition,situational motivation,and information seeking behavior.[Result/Conclusion] The impact and action path of the users' individual health risk perception status had different differences on both problem-solving related variables and health risk information seeking behavior of COVID-19.

[Purpose/Significance] Online health information seeking has become one of the most popular everyday life information behaviors.It is significance to understand consumers' online health information seeking needs and perceived barriers to improve the quality of health information service.[Method/Process] In this paper,34 participants were recruited for in-depth interviews.They were asked to recall and describe an impressive recent experience of online health information seeking.The critical incident technique was used to code and analyze the topics of the interview data.[Result/Conclusion] Based on the analysis results,three topics were obtained,namely,consumers' seeking channel needs,health information needs and perceived seeking barriers.Eight sub-topics of seeking channel needs,nine sub-topics of information needs and eight sub-topics of perceived barriers were further extracted.This study provides a reference for understanding consumers' everyday life health information seeking behavior and improving health information service.

[Purpose/Significance] This article explores the evolutionary regularity and dynamic mechanism of online health community users' continuous participation behavior,with a view to providing a useful reference for the online health community platform to promote user continuous participation and its service optimization.[Method/Process] Based on the social support theory and motivation theory,the costs and benefits of online health community users' continuous participation behavior were defined,the game benefit matrix was established,the online health community users' continuous participation behavior evolutionary game model was constructed,and the dynamic evolution process and equilibrium stability strategy were analyzed.[Result/Conclusion] The results showed that the continuous participation behavior strategy of online health community users was closely related to the cost,reputation benefits and social support benefits paid by other users,and when the cost and benefits paid by users change,the continuous participation behavior of online health community users would show different evolutionary regularity.

[Purpose/Significance] Affected by COVID-19,traditional offline academic conferences are gradually changing to online mode to meet the needs of academic knowledge exchange in special periods.This paper takes the participants of online academic conferences as the research object,and explores the influencing factors of their willingness to participate in knowledge exchange.The relevant research helps to better understand the attitude of researchers towards the online academic conference mode,and also provides some inspiration and thinking for the construction of future academic knowledge exchange mode.[Method/Process] Based on UTAUT model,this study constructed a knowledge exchange participating-willingness model of online academic conferences including performance expectation,effort expectation,social impact,service quality and communication context.379 valid questionnaires were obtained by questionnaire survey,and the variables and hypotheses in the research model were tested and analyzed by SPSS 24.0 and AMOS 24.0.[Result/Conclusion] The results showed that performance expectation and effort expectation had a significant positive impact on researchers' willingness to participate in the online academic conference knowledge exchange;social impact and service quality had no significant impact on the willingness to participate in online academic conference knowledge exchange;communication context had a significant negative impact on researchers' willingness to participate in the online academic conference knowledge exchange,which was due to the disadvantages of the "cloud" discussion mode in interpersonal experience,which shows that there were still some difficulties in academic knowledge exchange to complete the online transformation.

[Purpose/Significance] At present,the research on the reproduction of the academic pedigree of scientists is still in the exploratory stage,and only a small amount of literature attempts to analyze its influencing factors from the perspective of the external environment.This research has constructed a model of the factors affecting the reproduction of the academic genealogy of scientists.It will help to enrich the research of the academic genealogy of scientists,and also provide guidance for the cultivation and improvement of scientists' academic genealogy reproduction ability.[Method/Process] Firstly,we took the academic genealogy of Yan Yimin,an information scientist,as the research object.Then,the factors affecting the reproduction of the scientists' academic genealogy were preliminarily analyzed by grounded theory method,resulting in 114 initial concepts,36 categories,and 5 main categories.[Results/Conclusion] The influencing factors included internal factors and external factors.Mentor factors,student factors,and the construction of academic genealogy constituted the internal factors.Academic environment and scientific research opportunities consisted of external factors.

[Purpose/Significance] The relationship between entities can be regarded as the basic carrier of knowledge.The extraction of entity relationship from unstructured data is the main problem to be solved in the development of knowledge.[Method/Process] From the perspective of the research object of unstructured text,entity relationship was extracted by using the Bert-BiLSTM-ATT model.[Result/Conclusion] In the experiment of entity relationship extraction based on small sample cultural relic information resources,it could discover the unmarked time and location relationship to a certain extent,which provided a practical and effective data acquisition method for the related research of knowledge discovery in cultural relics information resources.

[Purpose/Significance] Academic blogs,academic WeChat public accounts,Academic APPs and other new academic media support a variety of knowledge social behaviors such as sharing,interaction,and dissemination of users.Revealing the mechanism of the social reading behavior of academic new media users helps to discover the psychological motivation of behavior and optimize the quality of knowledge social services,so as to promote accurate push and reading promotion.[Method/Process] On the basis of the process of social reading behavior of academic new media users,generation mechanism models were constructed from social learning theory(SLT) and self-determination theory(SDT).The association between two mechanisms was also analyzed.[Result/Conclusion] The motive psychology of user social reading behavior was clarified.SLT's triadic reciprocal determinism,self-efficacy,learning from observation,self-regulation,behavior reinforcement and SDT's autonomy need,competence need and ownership need jointly drive user' knowledge social behavior.

[Purpose/Significance] Intelligence system of public health emergency is the information support for preventing and responding the public health risk.Discussing the optimization of intelligence system of public health emergency can be utilized to improve the level for preventing and controlling public health risk,and further achieve the modernization of emergency management.[Method/Process] Based on the analytical framework of risk chain,this paper analyzed the intelligence system of public health emergency in two dimensions of "preparedness" and "response",and explained the path to optimize the intelligence system of public health emergency from three elements,including surveillance,communication and decision-making.[Result/Conclusion] The intelligence system of public health emergency should focus on the surveillance of potential public health risk,including public health-related and environment-related information.Also,intelligence communication should be enhanced for evidence-based decision-making by scientific intelligence,which enables to integrate "intelligence" and "decision-making" in the risk governance.

[Purpose/Significance] To explore the impact of the big data environment and the characteristics of strategic emerging industries on the company's competitive intelligence work,and to provide reference for related companies to carry out competitor research in the future.[Method/Process] This article analyzed the research process and system of strategic emerging companies' competitors in the big data environment,used Vensim PLE software to construct the dynamic model of the strategic emerging company's competitors research system in the big data environment and conducts simulations.[Results/Conclusions] The simulation model and results simulated the competitor research process of strategic emerging companies in a big data environment,and can provide a theoretical basis for strategic emerging companies to formulate effective competitor analysis strategies.

[Purpose/Significance] Technology convergence has become a mainstream innovation trend,stimulating the emergence of new and emerging technologies,and understanding and detecting this convergence process in a timely manner is important for formulating development strategies.[Method/Process] The paper identified the life cycle of technology development through the growth curve on the patent data,comprehensively considered the knowledge flow between the main and subsidiary IPC of the patent and during the citation process to establish a directed knowledge flow network,and identified the key integration through centrality indicators on a micro basis,Map IPC numbers to different industries through the IPC-ISIC standard,used Markov models to predict the proportion of industries in the future and analyzed the laws of industry integration at different stages.[Result/Conclusion] The graphene field was currently in a mature stage,and its key technology fusion components in the budding,growth and maturity stages have been discovered.The proportion of each industry in the graphene field at the end of the maturity period was predicted and one of the industries was used as an example to reveal its The law of integration at different stages.

[Purpose/Significance] In order to help digital content service providers formulate supply chain strategy,optimize digital content service efficiency and win the market.[Method/Process] Based on the analysis of digital content cost and demand of digital content service providers,this paper first introduced the network externality,constructed the mathematical model of digital content dynamic replenishment decision,then analyzed the influence of network externality and cost on the supply chain strategy of digital content service providers by operations research methods,and finally put forward the supply chain strategic decision model of digital content service providers.[Result/Conclusion] According to the cost and network externalities,the model can dynamically generate parameters which represent different strategy,and assist digital content service providers in selecting supply chain strategies.

[Purpose/Significance] The guidance of online public opinion dissemination is of great significance to reduce the risk of online public opinion,improve the effectiveness of online social governance and the ability of public health emergencies prevention and control.[Method/Process] Based on the perspective of "information users",starting from the three roles of online public opinion reporting,dissemination and guidance in public health emergencies,this paper analyzed the profit and loss of the three types of game players in the process of online public opinion dissemination,including information producers,information transmitters and information disintegrators,and analyzed the stability strategy by establishing a "tripartite evolutionary game" model.Finally,taking the "Shuanghuanglian incident" as an example,a numerical analysis was performed to analyze the evolutionary game strategies of the three types of agents.[Result/Conclusion] The experimental results showed that the key factors affecting the dissemination of online public opinion in public health emergencies were information quality and punishment intensity of information producers.Based on this,countermeasures and suggestions for the management of online public opinion were proposed.

[Purpose/Significance] Weibo communication effect evaluation can help enterprises to more accurately understand the communication of their brands,events,public opinions and other information on social platforms.However,due to the negative influence of water army and fake fans,it is difficult to measure the true communication effect of microblog.[Method/Process] Based on the interaction behavior of Weibo users,this paper proposed the algorithm of user interaction influence degree on the basis of interaction intensity,and introduced the concepts of topic communication depth,width,rate,efficiency and user communication contribution degree,so as to evaluate the communication effect of enterprise microblog in a relatively complete way.[Results/Conclusion] Simulation results showed that the interaction influence degree algorithm was in good agreement with the real situation,which can effectively eliminate the influence of false interactive data.The verification results of the official microblog of "Huawei Mobile" showed that the above methods can conduct in-depth analysis of the communication effect from multiple dimensions,and the analysis framework was reasonable and effective.The analysis results can also provide suggestions for the improvement of brand communication and social marketing of enterprises.

[Purpose/Significance] This paper aims to explore the influence mode and effect of group behavior on individual users in mobile social networks.[Methods/Process] The SEIR model was optimized and improved to increase the influence mechanism of group size,conformity effect and social reinforcement effect,and the public opinion communication mechanism of the four groups in the social network was analyzed dynamically.[Results/Conclusion] The simulation analysis was carried out based on scale-free network.The simulation results showed that the large-scale community public opinion communication would have intermittent impact on the small-scale community,the crowd effect had a significant impact on the depth and breadth of information dissemination,and social reinforcement effect would increase the probability of information dissemination in a certain range.

[Purpose/Significance] The spread of Internet rumors will disturb public order and cause public anxiety and panic.It is of great theoretical and practical significance for the harmonious development of network ecological environment to study the supervision strategy of network rumor.[Method/Process] Based on the interaction of government,media and Internet users,this paper constructed a game model of Internet rumor supervision.It used the method of system dynamics to simulate the evolution of the game model.This paper used sensitivity analysis to explore the influence of exogenous variables on game strategy.[Result/Conclusion] The final evolution results of four different local equilibrium strategies were (1,1) strategies.The increase of the participation of Internet users' supervision can effectively improve the evolution speed of game strategy.The change of initial supervision cost has the greatest influence on government game strategy,and the change of government reward has the greatest influence on media game strategy.