［Purpose/significance］ In order to clarify the overall situation,research trend and research hotspots in the field of cluster information security abroad.［Method/process］ Based on the relevant literature published by foreign scholars,this study sorted out,analyzed,and classified the relevant research in this field.［Result/conclusion］ The research summarized the concept of cluster information attack.It summarizes these as three main forms:information cluster,confusion between true and false information and information subject accumulation.Then,the study further explores the new forms of artificial intelligence-generated content (AIGC) technology applied to cluster-based information attacks,such as cluster-based deep forgeries,natural language fabrications,and emotional demagoguery,which are more targeted and destructive,posing a higher challenge to the existing information security defense system.Finally,this paper summarizes the detection and protection strategy of AIGC technology applied to cluster attacks,analyzes the current management pain points of cluster information security problems,and proposes future research trends.

［Purpose/significance］ Technological catalysis highlights the production characteristics of large scale,low homogenization threshold,and frequent use of generative tools in information,which are expected to serve as weaponization tools under the promotion of unrestricted spatial agglomeration.It is particularly necessary to explore preventive measures and strategies from the perspective of information perception.［Method/process］ Identify the phenomenon of large-scale homogeneous generative information aggregation from four perspectives:information flow conditions,homogenization attributes,generative tool roots,and agglomeration evolution.Based on this,establish a targeted information perception process and integrate it into a two-step method design to achieve research objectives.Exploring the strategy for sustained construction of information perception capabilities based on empirical results and their underlying reasons,starting from the characteristics of phenomena.［Result/conclusion］ Research has found that the method has good performance and can achieve the representation of intelligence content in terms of information aggregation scale,aggregation homogenization degree,aggregation characteristics,and aggregation threat dynamics warning,thereby providing support for the institution of security perception nerve endings.［Limitations］ The information perception results only focus on threat detection,lacking a focus on the causes of phenomena.In the future,in-depth theoretical analysis can be conducted to more comprehensively reflect the underlying mechanisms beneath the surface.

［Purpose/significance］ One of the background products of AIGC,cluster based deepfake information,is prone to severe real-world harm due to its subjective malignancy,difficulty in detection,identification,and control,making its attack information perception particularly important.［Method/process］ Design a deepfake information recognition scheme based on the GJK algorithm,establish a cluster feature analysis model based on Manhattan distance,and provide a structured intelligence system using portrait technology as a framework to achieve perception method design under attack intent determination.To verify the effectiveness of the method,it is incorporated into a simulation process supported by wargame deduction and using CMO software as a tool,and executed based on environmental configuration,interactive scenario unit assignment,and application matching.［Result/conclusion］ The perception results show that the method design can effectively achieve deepfake information recognition,cluster feature analysis,and intelligence portrait extraction functions.To maximize the effectiveness of research,we will explore strategies for building information perception capabilities based on deduction results.［Innovations/limitations］ Innovation lies in providing a complete set of clusters based deepfake information attack information perception methods.The shortcoming is that the content focuses on methodological research and does not specifically discuss the theoretical issues involved.

［Purpose/significance］ To comprehend the new changes in the knowledge value chain under the influence of digital intelligence and clarify the theoretical model and essential features of the digital intelligence knowledge value chain.The aim is also to enhance the organization’s knowledge value-added capabilities and competitiveness.［Method/process］ First,the connotation of digital intelligence knowledge value chain is summarized based on the research experience of the academia on knowledge value chain,digital intelligence empowerment,knowledge chain and value chain,and the theoretical model and core features of digitalized knowledge value chain are proposed.Second,the operational mechanisms and intrinsic and extrinsic motivations of the digital intelligence knowledge value chain are summarized to analyze the necessity of digitalized transformation of the knowledge value chain.［Result/conclusion］ Empowered by digital intelligence,the knowledge value chain has new characteristics and operating mechanisms,and the digital transformation of the knowledge value chain has become the key to improving the competitive strength of organizations.

［Purpose/significance］ The paper construct a value appreciation backtracking management model for reusing scientific data,in order to reflux,organize,evaluate,accumulate,and use various value appreciation achievements of reusing scientific data.［Method/process］ Based on the perspective of reinforcement theory,we define the management elements of value appreciation backtracking for reusing scientific data,including reinforcement subjects,reinforcement objects,reinforcement behaviors,reinforcer,and scientific data.By sorting out the functions and relationships of each element,explain the mechanism of value appreciation backtracking for reusing data from the perspective of reinforcement theory.Then a four layers reusing data value appreciation backtracking management model was constructed,and the institutional layer,rule layer,business layer,analysis application layer of the model were interpreted layer by layer.［Result/conclusion］ The value appreciation backtracking mechanism and management model of reusing scientific data obtained from the perspective of reinforcement theory can supplement and improve the subsequent management process of scientific data sharing,thereby achieving in-depth discovery and continuous release of the implicit value of scientific data.

［Purpose/significance］ At present,combination of Altmetrics and citation metrics is increasingly becoming a trend of academic influence evaluation on scholars.However,academic potential influence is usually overlooked and interaction between scholars involved in the process of academic dissemination has not yet attracted enough attention in the existing evaluating system.Therefore,it is difficult to meet the needs of comprehensive and accurate evaluation on scholars in the complex information environment.［Method/process］ The Defleur’s Interaction Process Model in the communication science and A-FAST Model in the highly potential talents evaluation theory inspire the research on multidimensional evaluation model of scholars’ academic influence in the perspective of interaction.On the basis of the theories,this paper complements the dimension of potential influence besides the dimensions of professional influence,disseminative influence and societal influence,constructs the multidimensional evaluation model of scholars’ academic influence as well as specific indication system of the four dimensions,and Identifies the applicable scenarios.［Result/conclusion］ According to the constructed multidimensional evaluation model of scholars’ academic influence as well as indication system,evaluating application on the Chinese scholars’ academic influence in the field of graphene is performed and the effectiveness and feasibility of the evaluation model is verified.［Limitations］ Limitations on both the single evaluated base data and the simplified verified scenarios remain to be promoted in the future research.

［Purpose/significance］ It is an important task that scholars reasonably select academic representative works,especially for general scholars,under the background of “vigorously promoting the evaluation of outstanding achievements and representative works”.［Method/process］ After elaborating the theoretical basis of model construction,this paper constructs a FIS model suitable for the selection of representative works of general scholars from three aspects:journal impact factor,immediately cited frequency and context similarity.Then,this paper makes an empirical study on the selection of academic representative works by using the academic papers of the winners of natural science excellent papers in Shaanxi Province.In addition,the comparative analysis is carried out from two aspects of internal and external comparison.［Result/conclusion］ The results proved that the model constructed in this paper has certain comparative advantages and practical significance.The FIS model can better select representative works,and it can help promote academic prosperity and progress to a certain extent.

［Purpose/significance］ The measurement of patent influence is of great significance to evaluate the value of patent technology and promote the cultivation and transformation of core patents.This study enriches and improves the evaluation of patent cumulative influence.［Method/process］ Based on the whole generation of patent citation network,the patent quality was integrated to construct the patent cumulative influence evaluation model,and the validity of the model was tested by case analysis method.Based on this evaluation model,Wilcoxon symbolic rank test was used to analyze the difference of patent cumulative influence between the total cited perspective and the other cited perspective.［Result/conclusion］ The results show that the evaluation model considers both the “quantity” and “quality” of cited patents,which can further distinguish patents with the same influence in the previous measurement,and can also give potential influence to the quality of zero-cited patents.Compared with the perspective of total citation,the cumulative influence of patents measured from the perspective of other citation is more consistent with the real situation.

［Purpose/significance］ To promote the enhancement of China’s scientific and technological innovation (STI) efficiency and the rational layout and healthy development of the region in the digital era,we try to study the efficiency of China’s provincial STI from the perspective of digital capital,and further optimize the investment of digital capital and the STI environment in each province.［Method/process］ By sorting out and refining the connotation of the broad concept of digital capital,the evaluation index system of China’s STI efficiency under the investment of digital capital is constructed.The three-stage DEA method and multi-source authoritative data are used to conduct empirical evaluation and analysis of the provincial STI efficiency.With the aid of the SFA model to explore the external influencing factors of the utilization rate of digital capital,according to adjusting the changes of provincial STI efficiency before and after the influence of environmental effects,managerial inefficiencies and statistical noise,analyze the regional differences and influence mechanism of provincial STI efficiency.［Result/conclusion］ The study shows that:there is still much room for improvement in the efficiency of STI under digital capital investment in China,mainly due to the lack of scale efficiency;there are problems of improper production structure and scale and underutilization of digital capital in provinces that have not achieved optimal technical efficiency;most provinces provide a good external environment for the development of STI under digital capital investment.It is suggested that the provinces that have not achieved optimal technical efficiency should optimize the investment of digital capital as the basis and effectively promote the effective implementation of China’s provincial STI development strategy by maximizing the differentiated development of location advantages.

［Purpose/significance］ Analysis of the evolution of patent technology topics can reveal the technological development trajectory in the subject area,and how to accurately mine it is crucial for enterprises to improve their technological innovation capability.［Method/process］ Along the research line of “technology evolution theoretical framework”→“technology life cycle division”→“technology topic extraction”→“technology topic evolution analysis”,taking the self-constructed database as the data source.Firstly，we divide the technology life cycle based on the Logistics model,and then extract the technology topics by using the LDA topic model and PCA+K-means model respectively,and then analyze the evolution of technology topics from the dimensions of technology topic intensity and technology topic content.Finally,this paper takes the video image processing technology field as an example to verify the validity and applicability of the model.［Result/conclusion］ The research results show that the intelligent model constructed in this paper can mine the popular and emerging topics in the subject area,and the analysis method has practical application value for recognizing and tracing the evolution law of technical topics and then clarifying the development vein of key technologies.

［Purpose/significance］ From the perspective of symbiosis,we explore the process and mechanism of knowledge co-creation evolution in online health community,which provides reference value for the deepening of knowledge co-creation theoretical research and the optimization of online health services.［Method/process］ Dissect the symbiotic evolution connotation of knowledge co-creation in online health communities,construct a Logistic-based equation model,simulate experiments on the effect of knowledge co-creation in online health communities under the invasion of different interaction modes,growth rates,saturation amounts and off-topic information actors,and introduce real online health community data as a source of data to conduct an example analysis.［Result/conclusion］ Based on the analysis of simulation experiments,online health community knowledge co-creation has seven interaction modes,growth rate and saturation amount have different influence paths,and the intrusion of off-topic information actors has different degrees of impediments to knowledge co-creation;the actual data proves that online health community knowledge co-creation has a symbiotic effect and reveals the real interaction evolution process.［Limitations］ The study only reveals the evolution process of knowledge co-creation from the overall macro perspective,and then reveals the evolution mechanism by comparing the knowledge content of different regions in online health communities.

［Purpose/significance］ Revealing emerging research topics with research and development potential in the field of artificial intelligence from conference papers can provide a path to capture the dynamics and future trends.［Method/process］ Establish a dataset for the literature of top international conferences in the field of artificial intelligence,apply the BERTopic model to detect topics,and utilize novelty,growth,and topic strength to comprehensively calculate the emerging potential of each research topic,and evaluate the emerging topics in the field.［Result/conclusion］ The overall development of the field of artificial intelligence is highly dynamic,and the comprehensive calculation considers “Object Detection”,“Federated Continual Learning”,and “Gradient Boosting” to be the most potential emerging topics at present.“3D point cloud”,“audio-vision crossover”,and “AI privacy protection” are promising emerging topics in the future.The research framework proposed in this paper is capable of discovering more current,accurate and interpretable findings with practical feasibility.［Limitations］ we do not carry out comparative research and comprehensive analysis of different neural network language models,and does not have sufficient refinement of data with different focuses within the dataset.In the future,a combination of different models can be utilized to reveal more systematic,targeted,and directional findings.

［Purpose/significance］ This paper aims to analyze the phenomenon of interdisciplinary knowledge flow,measure the characteristics of knowledge at different flow stages,and classify the knowledge.［Method/process］ The interdisciplinary knowledge flow is divided into the stages of knowledge inflow,concept innovation,and knowledge outflow.Then,through interdisciplinary analysis,knowledge emergence analysis,interdisciplinary citation analysis,and conceptual compatibility analysis,the characteristics of knowledge at different stages are measured.Based on the differences in characteristics,the knowledge is classified into four types:borrowing type,assimilation type,innovation type,and differentiation type.［Result/conclusion］ By understanding the characteristics of interdisciplinary knowledge at different stages and classifying it from the perspective of knowledge flow,a better understanding of the conceptual differences of interdisciplinary knowledge in different disciplines can be achieved,and its development direction can be predicted.

［Purpose/significance］ Medical named entity recognition is a critical step for medical and healthcare knowledge mining and knowledge organization.The semantic associations between multimodal data are mined to improve the medical entity recognition effect,which provides support for domain knowledge complementation and knowledge reasoning.［Method/process］ In this paper,we propose a multimodal Chinese medical named entity recognition model BAF-MNER (Bilinear Attention Fusion-Multimodal Named Entity Recognition) based on bilinear attention fusion mechanism.The model first learns semantic features from multimodal medical data through visual and text encoders;next a bilinear attention network is utilized to achieve cross-modal semantic interaction between image and text,introducing a gating mechanism to filter the visual noise;then fusing the visual and text features based on the attention mechanism to construct the multimodal feature representations,and adding a batch normalization layer to optimize the deep neural network;finally,inputting the multimodal feature vector into a CRF layer to decode to obtain the predicted labels.［Result/conclusion］ The proposed model can effectively improve the Chinese medical named entity recognition,increasing the F1 value on multimodal medical dataset by 4.07% compared with the unimodal baseline model,and 1.65% compared with the multimodal baseline model;the experiments on multimodal public dataset indicate an excellent generalization of our model.

［Purpose/significance］ Based on the perspective of “paper-patent” correlation,this paper identifies emerging terms that “lead the development trend of future technology in the innovation chain from paper to patent” through the way of emerging technology extraction and quantitative research.［Method/process］ This research innovatively adopts Termolator algorithm and GPT prompt learning to extract new terms.This method explores the application effect of GPT prompt learning in term extraction through comparative experiments,and significantly improves the accuracy and recall rate of term extraction.Furthermore,Minibatch Kmeans++ algorithm is used to cluster the results of term recognition,forming technical topics,and these emerging technical topics are identified and classified by multidimensional index quantitative analysis method.［Result/conclusion］ This paper divides the emerging technical terms into hot,cutting-edge,applied and potential emerging terms,and realizes the effective recognition and classification of the topics of technical terms.The results show that this method can effectively reveal the emerging technologies that lead the development trend of future technologies in the field of large model research and provide a new way for the identification of emerging technology terms.［Limitations］ This study has certain limitations in the vectorization representation of technical terms and the determination of the threshold of emerging technology topic recognition indicators which deserve further study.

［Purpose/significance］ The purpose of this study is to analyze the language content of Artificial Intelligence Generative Content (AIGC) in the context of online healthcare.The aim is to provide decision support for doctors using AIGC to respond to patient queries and for patients engaging in self-help diagnosis.Additionally,the study aims to offer reference insights for online healthcare platforms to effectively utilize AIGC and enhance service quality.［Method/process］ 1264 doctor-patient dialogue data were collected from the online medical platform,and patient consultation questions were input into ChatGPT to generate response text.The similarities and differences between the real response from doctors and the response generated by AIGC were analyzed in terms of text,vocabulary,semantics and emotion.TF-IDF model was used for text similarity analysis,Doc2Vec model was used for semantic similarity analysis,xmnlp toolkit was used for sentiment analysis,and BERTopic model was used for topic difference analysis.［Result/conclusion］ The average text similarity and semantic similarity of AIGC and doctor’s true response texts were 0.09 and 0.81,respectively,and the median score of affective attitudes was 0.48 and 0.44,respectively,with no statistical significance.There are differences between AIGC and doctor’s true response texts in space,level and similarity among subjects,and differences in content and representation effectiveness within subjects.The results show that AIGC has certain language understanding and situational reasoning abilities,which can provide some decision support for doctors and patients in online medical situations,but the accuracy,coherence and readability of language expression still need to be improved.