[Purpose/Significance] The appearance of ChatGPT led the rapid penetration of generative AI into the various walks of life and reshaped the industrial form.In this context, information science will inevitably be greatly affected.This paper discusses the research and practice trend of information science driven by generative AI, aiming at providing reference for information science to better meet the opportunities and challenges brought by AI reform. [Method/Process] Firstly, the influence of generative AI on information science research was discussed from three aspects: research problems, data sources and research paradigms.Secondly, this paper deeply analyzed the changes that generative AI would bring to information practice from four aspects: comprehensive knowledge service, academic information service, decision-making information service and social information service.Finally, combined with the limitations of ChatGPT, the problems that information science should consider when integrating generative AI and the contributions that information science can make were discussed. [Results/Conclusion] Generative AI is likely to have a great impact on information science, but if handled properly, it can promote the promotion of information science and innovate the information service model.Information science should actively embrace the new generation of artificial intelligence and explore the path of cross-integration with generative AI, but at the same time it needs to maintain an objective and prudent attitude.

[Purpose/Significance] In this era of data intelligence, data intelligence is a new term and a major frontier topic in the field of big data and artificial intelligence integration.Therefore, it is of great theoretical significance and academic value to accurately answer the most fundamental theoretical problem for data intelligence, namely what is data intelligence. [Method/Process] By using the methods of theoretical speculation and logical reasoning, based on a summary on existing explanations on data intelligence concept, this paper extracted the common understandings of data intelligence concept, and proposed a definition of data intelligence.Based on these, this paper proposed the elements of data intelligence, and explained the multiple meanings of data intelligence. [Results/Conclusion] Results of this study show that, data intelligence refers to a ability that is the analysis and extraction of useful information for management from data for a real scenario in order to service and support management.The four core elements of data intelligence are scenario, data, analysis and management.Meanwhile, data intelligence can be regarded as a thinking, product, process, ability, or tool and technology.

[Purpose/Significance] Global industrial data governance is the foundation for realizing the digital transformation of the industry, and high-level industrial data governance capabilities can effectively release data efficiency and enhance data value, thereby empowering new formats and models for industrial development. [Method/Process] This paper constructed a three-dimensional framework of"institution-market-technology", deconstructed the content of in-depth interviews into 9 subdivision indicators, and used a combination of qualitative comparative analysis(QCA)and necessary condition analysis(NCA)to study the improvement path of capability of industrial data governance under the background of digital intelligence from a configurational perspective. [Results/Conclusion] The level distribution of global industrial data governance capabilities of provincial governments in China shows obvious regional differences and agglomeration radiation effects.The intensity of capital investment is the key bottleneck of the Industrial datagovernance capabilities under the background of digital intelligence drive.There are five ways to improve the industrial data governance capabilities.They provide a reference for the"road to the same destination"path for the regions with similar basis of other elements but low ability of global industrial data governance.

[Purpose/Significance] At present, friend recommendation is widely used in social network platforms.High quality friend recommendation can not only meet users' needs for making friends and help users obtain more information resources, but also improve the stickiness between users and social networks. [Method/Process] Through combing, summarizing and summarizing the relevant literature, this paper defined the concept of friend recommendation, combed the process of friend recommendation in social networks, summarized the methods of friend recommendation, and summarized the shortcomings of existing relevant research and the focus of future research. [Results/Conclusion] The research found that friend recommendation in social networks helps users find other users they are interested in based on the records left by users on social network platforms.The existing friend recommendation research has problems such as cold start, data sparsity, dynamic change information cocoon room and privacy problems.In the future, friend recommendation can be studied from the aspects of multidimensional recommendation, providing explanations for results of friend recommendation, setting user feedback mechanism and improving the scalability of the recommendation system.

[Purpose/Significance] To explore the influencing mechanism of misinformation identification intention of social media users is helpful to improve the knowledge system of misinformation governance in China and provide a reference for perfecting the social correction mechanism of misinformation. [Method/Process] Based on Planned Behavior Theory, this paper introduced two variables: digital literacy and risk perception to construct a model of social media users' misinformation identification intention, and analyzed the influence mechanism of digital literacy, risk perception, recognition attitude, subjective norm and perceived behavioral control on the misinformation identification intention.Based on 371 sample data, the structural equation model was used for the empirical study. [Results/Conclusion] The results showed that digital literacy and perceived behavioral control positively affected the misinformation identification intention of social media users, while subjective norms and recognition attitude had no significant effect on recognition intention.Perceived behavioral control not only directly affects recognition attitude, but also mediates the relationship between digital literacy and recognition intention.Digital literacy not only directly affects the identification intention, but also indirectly through risk perception and perceived behavior control, which is the most important factor driving the identification intention of social media users with misinformation.

[Purpose/Significance] Excessive online health information-seeking behavior cause health anxiety and further induce cyberchondria, which can, in turn, exacerbate health anxiety by trapping people in an"information cocoon".The current study explores the causes of cyberchondria through the perspective of communication in order to provide insights for"breaking the cocoon". [Methodology/Process] This research constructed a structural model based on self-determination theory from two dimensions of exogenous and endogenous factors, and analyzed the data through the partial least squares path analysis method of SmartPLS. [Results/Conclusion] The results of the study show: ①Cyberchondria leads to the health-related information cocoon; ②positive metacognition, health self-efficacy, online health information credibility, and information overload positively affect cyberchondria and further lead to information cocoon; ③there was no significant association between subjective norms and cyberchondria.The current study expands the scope of research on cyberchondria and online health communication, and provides a reference for practical governance.

[Purpose/Significance] Named entity Recognition(NER)of Wechat service information is an important research direction in the field of intelligent service of university libraries.At present, in the field of NER of Wechat service's tweets, there are insufficient problems such as entity name and feature information extraction in the field of the university library Wechat service. [Method/Process] For the above problems, this paper used the knowledge base and dictionary rules to expand the sentence semantic information by mining the wechat service tweet text in the field of university library service.Under the guidance of the experts, which constructed a corpus set of university library wechat service's NER of the ULICNER.The data set contained 36035 pieces of text corpus data, including 7 categories and 16 subcategories.Then, the model used BERT network to generate character vectors with semantic features from large-scale text, then obtained text features from the input text sequence through BiLSTM, and finally connected CRF to obtain the optimal output of sequence tags. [Result/Conclusion] The experimental results show that the model in this paper achieves good results on the constructed data set.The F1 value is 98.75%, the accuracy value is 98.59% and the recall rate is 98.91%.Finally, the model in this paper can be applied to the named entity recognition of university library information service.

[Purpose/Significance] The quality model provides unified terminology and guidance for quality assessment.However, the current quality models has some problems such as terminological?incongruency and poor applicability.Although the W3C and some standardization organizations have also introduced some data quality knowledge representation models, they are more versatile and not operable because they belong to the top-level ontology of quality.How to make these standardized data quality knowledge representation models useful in specific areas and reduce the threshold for application has become a challenge.Based on the above questions, it is important for this paper to study how to construct a standard, consistent, and applicable methodology for the representation of quality knowledge. [Method/Process] Firstly, this paper used the traceability method to sort out and summarize the DQV and other four quality models, extract the core elements common to the quality model and organize them into a data quality concept model, so as to solve the problem of terminology inconsistencies.Secondly, this paper proposed a set of standard and applicable quality model modeling methods that built a standard and applicable quality model from the actual business needs from the actual business requirements.Finally, taking the semantic accuracy requirement as a case, the corresponding quality model was constructed by using the modeling method proposed in this paper and the feasibility of the proposed method was verified. [Result/Conclusion] In this paper, a three-stage, six-step quality model modeling method is proposed; First, this article standardizes the terminology of the quality model.Second, the modeling approach proposed in this paper provides users with an idea of how to apply and extend existing top-level quality models and build quality models that are suitable for specific domains.Finally, this paper verifies the feasibility of the proposed method.

[Purpose/Significance] Exploring the impact of different types of public crisis information on the public's risk perception and protective behaviour is crucial to the adoption of targeted communication and mobilisation strategies by information publishers.This study investigates the effects of public crisis information sources, narratives and their interaction on public risk perception and protective behaviour from the public's perspective based on an Elaboration Likelihood Model, which is important for improving the effectiveness of information persuasion and public compliance with protective behaviour. [Method/Process] A 2*2 randomised controlled pre and post test experiment was designed to test this study. [Results/Conclusions] The results show that: (1)The narrative method of crisis information has a positively significant impact on the public's risk perception and protective behavior, and the data-based method is more convincing than the narrative method.(2)Crisis information sources only have a positively significant impact on public risk perception, and crisis information from official sources is more persuasive than that from unofficial sources.(3)The interaction of information sources and narrative style has a significant positive effect on public risk perception and protective behaviour.(4)Public risk perception plays a positive mediating role in the relationship between crisis information sources, narrative methods and protective behaviors.Accordingly, to promote the public's protective behavior, we propose to use narrative methods in a suitable way, build a cooperative matrix of sources, and guide risk perception reasonably.

[Purpose/Significance] Domestic data exchange platforms still slightly insufficient in industrial innovation incentives compared with foreign platforms.In this case, how to make platforms develop towards the target of stimulating industrial innovation has become a very important issue.However, the development of platforms is affected by many factors.Therefore, only by identifying these factors can the platform be better optimized. [Method/Process] On the basis of introducing the relevant concepts of data exchange and sorting out the literature, the author listed the mainstream data exchange platforms, and analyzed the exchange process of Guiyang data exchange platform.Then, taking the platform as an example, a theoretical model was constructed by applying grounded theory to analyze the factors that affected the platform operation. [Result/Conclusion] The research results show that there are four main factors affecting the operation of Guiyang data exchange platform, namely security guarantee, platform design, operation environment and data resources, followed by the explanation of the theoretical model constructed.

[Purpose/Significance] The analysis of online public opinion communication mechanism and find the influencing factors under blockchain environment can better ensure the identifiability of false information communication in social media, and play a positive role in realizing online public opinion governance and creating the ecological space of network public opinion. [Method/Process] Based on the information communication theory and information ecology theory, this study analyzed the communication mechanism of online public opinion in the blockchain environment, and constructed a model of influencing factors of online public opinion communication in the blockchain environment.Web crawler was used to obtain social media data of Minds platform to verify the proposed model and hypothesis. [Result/Conclusion] The results show that the popularity of blockchain environment, blockchain information preference, blockchain technology preference and user's influence are the important factors that affect the network public opinion communication under blockchain environment.Among them, blockchain environmental heat, blockchain information preference, blockchain technology preference and user's influence have positive effects on blockchain network public opinion communication.While image information uploading preference and video information uploading preference have positive effects on blockchain information preference, which is negatively correlated with blockchain information preference, blockchain information preference has positive effects on blockchain technology preference, blockchain information preference has positive effects on blockchain environmental heat, and the number of fans and users who follow the chain have positive effects on user's influence.This paper studies the theory and algorithm of the online public opinion communication process in the blockchain environment, and puts forward certain countermeasures and suggestions for managing the ecological space of network public opinion in the blockchain environment from the influencing factors of online public opinion communication.

[Purpose/Significance] Explore the rules of online public opinion audience participation behavior, and achieve accurate targeted guidance of online public opinion audience participation behavior. [Method/Process] Considering the two-level penetration and overlap of public opinion audience participation behavior, a binary state improved SIR model was built, and a targeted guidance model for online public opinion audience participation behavior was built.A targeted guidance decision algorithm was proposed, and simulation research was carried out. [Result/Conclusion] The initial target component and public opinion guidance degree have a certain impact on the initial audience density of each layer, but have little impact on the overall initial audience ratio.The change of the guidance rate has a significant impact on the inter layer target conversion, and the neutral layer plays a bridge role in the internal drive behavior of the public opinion audience.At the level of driven behavior of public opinion audience, neutral emotion is the optimal target for public opinion guidance.

[Purpose/Significance] Privacy permission request of mobile applications doesn't attract enough attention from users, instead making them burn out on privacy issues and unwilling to put effort into protecting personal information.Exploring the factors influencing users' privacy fatigue and designing a new privacy permission request interface is important to alleviate users' negative psychological state and motivate them to adopt privacy protection measures. [Method/Process] Based on multidimensional development theory, this study constructed the influence mechanism model of information privacy behaviors(information disclosure intention and privacy protection disengagement intention)from the dual perspectives of technical characteristics(interface design perception, permission justification)and individual differences(privacy protection self-efficacy, privacy knowledge).The Structural Equation Model(SEM)was used to verify the ration of model construction. [Result/Conclusion] Compared with the existing privacy permission request interface, users' overall satisfaction with the privacy permission request interface designed in this study is higher.A high level of interface design perception, privacy protection self-efficacy and privacy knowledge play significant roles in alleviating users' privacy fatigue.Privacy fatigue positively influence users' information disclosure intention and privacy protection disengagement intention, instead privacy concern negatively influence users' information disclosure intention and privacy protection disengagement intention.Privacy fatigue offers a new perspective to explain the'privacy paradox'phenomenon.

[Purpose/Significance] The identification of emerging technologies helps to track the development trend of technologies in time, find out the development paths of technologies, and capture the future development opportunities of the field earlier.However, on the one hand, emerging technologies have not yet formed a unified definition and characteristic attributes; on the other hand, with the deepening of research and the application of new technologies, the identification methods of emerging technologies are increasingly diversified and complicated.Therefore, it is urgent to sort out and summarize the definition attributes, identification methods and evolution methods of emerging technologies. [Method/Process] Through the investigation of relevant literatures at home and abroad, the concept and attribute characteristics of emerging technologies are firstly clarified, and similar concepts were discriminated.The following was a systematic review of emerging technology identification methods, focusing on quantitative analysis methods, and dividing them into scientometric analysis, text mining analysis and machine learning method, also revealed the new trend of the emerging technology identification method.In addition, new trends in emerging technology identification methods were revealed, as well as common methods for result validation. [Result/Conclusion] On the basis of revealing the advantages and disadvantages of various emerging technology identification methods, the common problems of existing methods are put forward, that is, the concept is not clear, the standard is not unified.Data source type selection is unbalanced and biased obviously.The"progressive"review research is the mainstream, and the"leap forward"prediction research is still to be developed.

[Purpose/Significance] As the international public opinion field continues to be treacherous with the spread of COVID-19, how to grasp the law of public opinion dissemination to enhance China's international communication power and discourse is a major issue facing theoretical research and practical exploration of intelligence in a new era. [Methodology/Process] Using data from WoS and CNKI databases as the source, and supplemented by literature analysis methods for topic mining and opinion refinement, the logic of"behavioral characteristics layer-dissemination mechanism layer-practical application layer"was followed to review the research themes of online public opinion dissemination on public health emergencies in the international public opinion arena and present future prospects. [Results/Conclusions] (1) This research is a systematic project of cross-fertilization and support of various disciplines, involving intelligence science, information science, system science, communication science and other disciplines; (2)Existing literature focuses on three themes: the behavior of dissemination groups, public opinion dissemination mechanism, and international communication power enhancement; (3)Future research can try to combine different disciplines across borders and focus on the perspectives of public opinion portrait shaping, communication law analysis, and communication strategy application.