Journal Information
Sponsor: China Society for Defense Science and Technology Information
                Institute No. 210,CNGC
International serial number: ISSN 1000-7490
Domestic serial number: CN 11-1762/G3
Mailing address: Box 10, Box 2413, Haidian District, Beijing
Postal code: 100089
Email: itapress@163.com/1587682149@qq.com
Tel: 010-68961793/68963306
WeChat public account: qbllysj
2025 Volume 48 Issue 5 
Published: 23 May 2025
  
  • Select all
    |
  • Qiu Junping, Zhang Tingyong, Xu Zhongyang
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Understanding the characteristic differences between false health information in User Generated Content (UGC) and Artificial Intelligence Generated Content (AIGC) is of great significance for improving the comprehensive governance system of false health information in the age of digital intelligence[Method/process] Based on 600 samples of false health information,the article used proceduralized grounded theory to identify the characteristics of false health information and explored the differences in the characteristics of false health information in UGC and AIGC using the contingency table chisquare test and analysis of variance (ANOVA).[Result/conclusion] False health information in both UGC and AIGC exhibits four core characteristics,nine primary characteristics,and 34 subcharacteristicsFor the same topics,four primary characteristics,including data bias and logical fallacies,are the main distinguishing features between UGC and AIGC false health information,while three primary characteristics,such as information confusion,are secondary featuresFurthermore,significant differences in subcharacteristics exist between UGC and AIGC within the same topic and primary characteristicFor different topics,in UGC false health information,information confusion is a prominent feature common to the topics of healthcare,food safety,and preventive care,while terminology manipulation,misleading statement,and false authority or others are prominent features of the topics of healthcare,food safety,and preventive care,respectivelyIn AIGC false health information,misleading statement is the notable characteristic in the topic of food safetyBased on these findings,the research proposes strategies to optimize the comprehensive governance system for false health information on the internet
  • Qiu Junping, Wang Yangbo, Xu Zhongyang
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Identifying and analyzing the key influencing factors of social media users’ false information verification behavior (FIVB) is of great significance for motivating users’ FIVB,improving the false information governance system,and promoting the healthy and sustainable development of the Internet information ecology[Method/process] The article constructed a model of factors influencing social media users’ FIVB by applying qualitative metasynthesis,risk information seeking and processing model,and media context theory,followed by identifying and analyzing the key influencing factors in it by using the DEMATEL method[Result/conclusion] The study shows that:① The model of influencing factors of social media users’ FIVB is mainly composed of 4 dimensions and 20 influencing factors,such as user factors,media factors,information factors,and contextual factors② The influencing factors interact with each other to form a complex network system③ The 11 influencing factors,such as user information literacy,media authority,and media popularity,are the key influencing factors of social media users’ FIVBBased on this,we propose optimization strategies to motivate users’ FIVB and improve the false information governance system

  • Sun Ying, Zhang Yu, Liu Shu
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Qian Xuesen’s academic thoughts on intelligence have guiding significance for the modernization and intelligent development of scientific and technological intelligence,emphasizing the core role of humanmachine integration in improving the quality of intelligence work[Method/process] The article outlines Qian Xuesen’s views on the positioning of intelligent technology,reviews how he integrated artificial intelligence into the scientific and technological system,and predicts its development trends in intelligence work[Result/conclusion] In the process of intelligent development,intelligence research should adapt to the technological changes,make full use of emerging technologies such as quantum computing and biocomputing to improve the efficiency and accuracy of information processingHowever,human wisdom or thinking activities are the leading factors in intelligence work and are key to determining the quality of intelligenceQian Xuesen’s thoughts on intelligent technology and intelligence emphasize the importance of humanmachine integration,proposing the construction of a humanmachine collaborative intelligent complex that complements each other’s strengths to achieve the scientification,precision,and efficiency of intelligence workTherefore,future intelligence work should pay more attention to interdisciplinary,crossdomain,and multiscenario integration and innovation to achieve the perfect combination of “quantitative intelligence” and “qualitative intelligence”,that is,the wisdom of great synthesis
  • Zhang Yufeng, Feng Yunhao, Lei Shuai
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] The dynamic information in the field of national defense science and technology is fundamental data for national defense science and technology information analysis and decisionmaking,the acquisition and utilization of which is an important practice for defense science and technology information work[Method/process] Based on the reality of fierce major power competition and rapid technological change,this paper analyzes the challenges faced by dynamic information work in the field of national defense science and technology,as well as the contradictions and problems existing in dynamic information tracking business of national defense science and technology,and puts forward the construction of dynamic information business model of national defense science and technology[Result/conclusion] The dynamic information business model of national defense science and technology can help clarify the nature of dynamic information tracking business of national defense science and technology,and support the dynamic information work in the field of national defense science and technology to carry out better
  • Yu Gaihong, Xie Jing, Zhang Zhixiong, Wang Ying, Zhang Qin, Liu Linlin
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] To cope with the new opportunities and challenges brought by intelligent information technology,support the upgrading of technology intelligence services,this paper conducts research on the theory,system framework and practical applications of intelligent information service[Method/process]A systematic investigation was conducted on the current theories and system platforms of intelligent information serviceBased on practical experience,a “3+1+N” intelligent information service framework based on DIKIW model was proposed,and the practical exploration of intelligent information services was carried out in specific science and technology fields[Result/conclusion] In the era of digital intelligence,intelligent services are increasingly emphasizing user scenariodriven and providing indepth services tailored to professional fieldsThe intelligent information service system framework proposed in this paper holds significant reference value for intelligently building domainspecific data foundations,knowledge foundations,and intelligence foundations,developing domainspecific intelligent intelligence service platforms,enhancing the intelligence level of scientific and technological intelligence work,and supporting research in the field of scientific and technological information
  • Wang Lun, Askar Hamdulla, Zhang Huaping, Geng Guotong, Lü Yanfei, Lei Shuai
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Large language models have made significant progress in the problem of generating text,providing new ideas for researchers in the field of open source intelligenceThis paper applies large language models to the field of open source intelligence to explore its ability to generate summaries in this field[Method/process] By constructing an open source intelligence summary generation dataset,using the Qwen 1514B model as the base model,and finetuning it using the LoRA technology,the Qwen 15OSINT model was obtained[Result/conclusion] The Qwen 15OSINT model was compared with the Qwen 1514B model in the open source intelligence summary generation taskThe results showed that the ROUGE1 index of the Qwen 15OSINT model increased by 03842,the ROUGE2 index increased by 02021,the ROUGEL index increased by 02494,and the BLEU index increased by 01348The finetuned model surpassed the base model in many evaluation indicators,demonstrating excellent summary generation capabilities[Limitations] Although the finetuned Qwen 15OSINT model has improved in performance,the construction of the highquality dataset it relies on still requires a lot of manual effortIn addition,the generalization ability of the model may be limited in different fields,which requires further research and exploration
  • Ju Zihan, Wang Yanfei, Bai Rujiang
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] In complex environment,open source defense data has different structures and complex semantics,so it is urgent to integrate multimodal semantic information timely and effectively to assist intelligence decisionmakingThis paper proposes a multimodal national defense data fusion model to enhance the capability of national defense intelligence analysis[Method/process] Firstly,the development of open source intelligence is sorted outSecondly,the complexity of defense data environment is analyzedThirdly,the key technologies of multimodal data fusion are summarized,and the fusion model is constructedFinally,combined with the multimodal defense data fusion case,the corresponding fusion enlightenment is proposed[Result/conclusion] In complex environment,multimodal defense data fusion should not only accurately grasp the granularity of semantic fusion,but also take into account the needs of intelligence tasks,and combine deep learning to achieve fast and accurate fusion of open source data
  • Wang Wentao, Tang Sijie, Qian Pengbo, Chen Yue, Jiang Weihan, Li Li
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Exploring the interaction mechanism between AIGC and irrational information behaviour from the research perspective of user experience helps to enrich and expand the research content in the field of user experience and information behaviour,and provides theoretical references for indepth analysis of the formation and feedback mechanism of users’ irrational information behaviour under the AIGC scenario[Method/process] Based on the Phenomenography to interpret the interview data on the interaction relationship between AIGC and irrational information behaviour under the perspective of user experience,six new findings on the interaction mechanism between AIGC and irrational information behaviour are summarized[Result/conclusion] From the perspective of user experience,AIGC can drive multiple types of irrational information behaviour,and the feedback effects of different irrational information behaviour on AIGC are somewhat different,and the results of the related research will help provide useful references for the optimization of content supply by AIGC,and the mitigation of irrational information behaviour by users
  • Wu Zhixiang, Zhou Xingyu, Zhu Xiaofeng
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Data security has become a major issue concerning national security,economic development,and social governanceEstablishing a data security risk early warning intelligence system at the national level is of critical significance in preventing major data security incidents[Method/processe] First,based on existing research and regulatory practices,the concept of national data security is defined,its risk characteristics are explored,to form a basic understanding of the early warning intelligence systemSecond,a theoretical framework for the national data security risk early warning intelligence system is designed,including four main functional modules:knowledge base and risk reasoning identification,comprehensive risk assessment,risk prediction and early warning,and intelligence service platformFinally,from the methodological perspective of “WuliShiliRenli” Approach,taking the data security of generative artificial intelligence as an example,the implementation path of the risk early warning intelligence system is planned[Result/conclusion] The national data security risk early warning intelligence system is essentially an organic integration formed by the deep collaboration of human expert wisdom,data and knowledge systems,and intelligent technological tools in the entire process of intelligence workIt can provide rapid and accurate intelligence services for national data security prevention and governance
  • Huang Longwei, Yu Liping
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] With the development of information technology and the increasing maturity of artificial intelligence technology,the construction conditions for multiattribute academic evaluation decision support systems are becoming increasingly completeHowever,research in related fields is still in its infancy[Method/process] On the basis of defining the concept of multiattribute academic evaluation decision support system,this article constructs a system architecture,including a big think tank,knowledge tank,evaluation index tank,method model tank,and rule tankThe key issues and optimization paths are discussed[Result/conclusion] The results show the uncertainty of decision results is a difficult problem in multiattribute academic evaluation decision support systemsIt is necessary to pay attention to the representativeness problem in expert database construction,the consensus problem in academic evaluation knowledge base,and the depth of artificial intelligence interventionIt is necessary to strengthen the research on the basic theoretical problems of multiattribute evaluation,use artificial intelligence to promote the construction of decision support systems,focus on evaluation quality diagnosis,and explore the construction of decision support meta systems,and attach importance to feasibility analysis

  • Huang Weixin, Bi Datian, Zhu Zimeng, Wang Chuanqing,
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Purpose/significance] In the era characterized by mobile interconnectivity,diverse social media platforms,and intelligent interaction,the information behavior of users across social media has become a forefront issue of interest across various disciplinesUnveiling its underlying mechanisms can enrich the theoretical framework of information behavior,deepen the understanding of multichannel information collaboration,and provide insights for platformbased information services[Method/process] Employing grounded theory,an exploratory analysis was conducted on 28 interview transcripts and 13 pieces of usergenerated content to construct the TouchpointDriveInhibitionModulation (TDIM) modelThis model encompasses 7 main categories,22 subcategories,50 basic categories,and 118 initial concepts[Result/conclusion] The touchpoints consist of situational triggers,conscious need triggers,and unconscious need triggersThe drive is the attraction of resources from other platforms,the inhibition is the demand anchoring of the original platform,and the modulation is influenced by information literacy and platform competitioncooperationUnder the influence of situational and need triggers,the information resources from other platforms facilitate crossplatform information behavior,while the demand anchoring of the original platform inhibits such behaviorAdditionally,users’ information literacy and the dynamics of competition and cooperation between platforms moderate these processes[Limitations] The revealed relationships await empirical verification and do not consider the crossdomain phenomenon between social media at home and abroad

  • Ran Congjing, Feng Ruojing, Li Wang
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] This study propose a method for identifying technological competitors by integrating patent texts,International Patent Classification (IPC),patent citation relationships,and patent counts,leveraging natural language processing and graph neural network technologiesThe ultimate objective is to achieve more accurate predictions of an enterprise’s technological competitors[Method/process] Firstly,BERT and OneHot methods are utilized to process patent texts and IPC classification information,respectively,generating text feature vectors and classification feature vectors,which are then concatenated to form fused vectorsSecondly,a patent citation network is constructed based on citation coupling and cocitation relationships among patentsThe Variational Graph Autoencoder (VGAE) model is applied to perform graph embedding learning on the patent information network formed by the fused vectors and patent citation network,obtaining lowdimensional embedding’s for each patentFinally,the embedding’s of all patents belonging to an enterprise are aggregated to form an enterprise vectorBy calculating the similarity values and reduceddimensional feature vectors among enterprises,combined with the number of patents held by each enterprise,a technological competition bubble chart is plotted to identify technological competitors[Result/conclusion] Taking BYD’s new energy vehicles as a case,this study ultimately identifies technological competitors such as Geely Automobile and Chery AutomobileThis method provides a reference for enterprises to formulate technological competition strategies[Limitations]This study does not fully consider the impact of time factors on the evolution of patent citation relationships and technological development trends,which represents a potential area for future research and improvement

  • Li Haocheng, Li Changling, Gao Feng, Liu Jiaqi, Wang Hao
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] From the perspective of “newness” and “emergence” of research themes,and the two stages of knowledge production and dissemination,identifying the emerging themes of disciplines is conducive to tracking the latest development of disciplines,and providing references for academic research and scientific and technological innovation[Method/process] The process of “emerging” research themes is similar to the process of object movement,which is driven by various forces such as citation driving force,innovation driving force,and attraction pulling force,etc,and starts to move,generating kinetic energy for production and disseminationBased on the two aspects of theme knowledge production and dissemination,and the two perspectives of “newness” and “emergence”,we constructed the theme emergence kinetic energy model,measured the degree of theme emergence by multiattribute synthesis,and identified the emerging themes of target disciplines[Result/conclusion] The research literature and its citation data in the sample intervals of 2018-2020 and 2021-2023 in the discipline of information resources management are selected as examples for empirical research,and it is found that the model can effectively identify the emerging themes of the discipline,and according to the changes of the theme emerging kinetic energy,it can be divided into the “novel emergence”,“stable growth”,and “stable growth” categoriesAccording to the changes in the emerging dynamics of the themes,the model can effectively identify the emerging themes and classify them into “novel and emergent”,“stable and growing”,and “declining and mature”,so as to realize the indepth analysis of the emerging themes

  • GuoYajun, Feng Siqian, Kou Xuying, Li Tianxiang, Huang Xin
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Purpose/significance] Library embodied knowledge service represent an innovative service model based on embodied cognition theory,designed to serve users’ cognitive processes,effectively meet their knowledge needs,and drive the continuous growth of libraries[Method/process] Through an investigation of practical cases and research findings of embodied knowledge services in libraries both domestically and internationally,it has been discovered that libraries have already embarked on embodied service explorations,including immersive reading,human libraries,sensoryintegrated space design,and maker spacesThese efforts exhibit characteristics such as the concretization of data acquisition,networked information interaction,clarification of knowledge internalization,and practical wisdom generation[Result/conclusion] A library embodied knowledge service system is constructed from four dimensions:foundational data,interactive space,knowledge situation,and practical wisdomLooking ahead,libraries should gain insights into users’ dynamic knowledge needs,strengthen librarians’ awareness of embodied services,provide multisensory spatial experiences,and create an environment for cocreation practices

  • Zhang Tao, Zhang Boya, Ma Haiqun
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] An indepth study of China’s data security policy diffusion and topic transfer can grasp the basic characteristics and laws of China’s data security policy diffusion,thus enhancing the scientific nature of data security policy formulation and implementation[Method/process] A comprehensive application of social network analysis and time series analysis methods is conducted to deeply explore the diffusion characteristics and theme shifts of the 258 data security policies issued by the central and local governments between 2014 and 2023[Result/conclusion] China’s data security policies mainly follow a topdown diffusion direction,with the most common diffusion pattern being from the central government to municipallevel local governmentsIn terms of thematic shifts,specialized data security policies integrate the core content of the Data Security Law into their respective regions or fields and implement detailed measures;however,embedded data security policies lacks innovative content and detailed measures tailored to their respective fields[Innovation/value] This paper proposes three optimization suggestions:continuously exploring the data security policy diffusion model of “pilot projects in local areas,nationwide promotion”,constructing a closedloop governance system for the full lifecycle of data security from the perspective of technological innovation,and building a trustworthy data space with the data security ecosystem as its foundationThese suggestions aim to provide references for improving China’s data security 
  • Zhang Jin’ao, Lu Xinyuan, Zhou Yifu, Meng Hua, Wang Xuelin
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] HumanAI trust is the basis and premise of high quality humanAI cooperationThe deepening of humanAI cooperation has further enriched the perspective characteristics of humanAI trustTherefore,deconstructing the perspective features of humanAI trust in the context of humanAI cooperation is of great significance to understanding and consolidating humanAI trust[Method/process] Based on ANN and Q methods,this study attempts to divide the dimensions of humanAI trust and empirically test itOn the one hand,based on the relevant literature in the past decade,the paper quantitatively clusters the formation perspective of human intelligence trust and qualitatively summarizes its characteristicsOn the other hand,the simulated scenario experiment of ChatGPT is taken as an example to empirically test the perspective characteristics of humanAI trust through ANN[Result/conclusion]Firstly,the quantitative analysis of Q method found that humanAI trust has dual perspectives of social entity and operational tool.The qualitative analysis further deconstructs the concrete characteristics of human intelligence trust from the two perspectivesSpecifically,humanAI trust from the perspective of social entity can be analyzed from the dimensions of perception,cognition and judgment,while humanAI trust from the perspective of operational tools is composed of the characteristics of certainty,stability and machine abilityIn addition,ANN’s empirical test results show that multiple features from different perspectives of humanAI trust have a significant impact on user satisfaction and continuous use behavior,which proves that the dimension division of humanAI trust in this study has good applicability

  • Hou Jianhua, Liu Hengyi, Li Danyang
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] In the context of the rapid development of science and technology,the application of new science and technology cannot be separated from the dissemination of science and technology,and technology diffusion is the communication process of science and technology,and is a necessary condition for the development and innovation of science and technologyPatents are important intellectual assets of enterprises,and the competition of individual patents among enterprises has turned to the competition of patent portfoliosHow to objectively and accurately measure the influence of patent portfolios on technology diffusion of enterprises is particularly important[Method/process] Due to the complex content of patent literature and the strong subjectivity and large granularity of traditional patent portfolio identification methods,this study integrated and formed a technology diffusion impact measurement model based on enterprise patent portfolio in order to more accurately measure the impact of patent portfolio on enterprise technology diffusionSample data of the new energy automobile industry were obtained from the incoPat patent platform,and the enterprise patent portfolio was obtained through Python language programming to achieve label extraction,vectorization and clustering in the technical field of patent textThe influence of the scope and diversity of the patent portfolio of sample enterprises on the diffusion of enterprise technology was calculated and analyzed[Result/conclusion] The empirical results show that patent portfolio plays a significant role in the diffusion of enterprise technology;The scope of patent portfolio has a positive effect on technology diffusionThe relationship between patent portfolio diversity and enterprise technology diffusion is inverted Ushaped within a certain range

  • Wan Jing, Ma Xiaoyue
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] This study aims to explore how the remention of online events in the digital age triggers and shapes public memory,revealing how social media,as a dynamic interactive platform,participates in the reproduction,dissemination,and evolution of memoryThis provides a new perspective for understanding the characteristics of collective memory and the mechanisms of information dissemination in the context of rementioned online events[Method/process] By constructing quantitative indicators of memory intensity,breadth,speed,and preference,computational models were employed to analyze the evolution patterns of public memory under the remention of online eventsIn addition,an indepth analysis of 384510 social media data points was conducted using actual cases[Results/conclusion] Stimulus events significantly triggered public recollection and emotional resonance with the original events,with different types of events affecting memory intensity,breadth,and speed in varying waysPublic memory preferences revealed thematic sensitivity and emotional tendencies in the process of memory reconstructionOnline event remention not only promoted the reproduction and communication of memory but also reshaped public cognition and emotion towards past eventsThe findings further deepen the understanding of the collective memory construction process in the digital age and provide a method for predicting the dynamics of public memory from the perspective of event clustersUsing public memory connections as a clue to capture the intermittent dissemination patterns of online event information;this study offers strategic references for optimizing the emergency management of online event information
  • Liu Yizhou, Huang Wei
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Mapping and fitting the risk behavior trajectory of online public opinion subjects in the information space of nonlinear risk superposition has become the main focus of public opinion supervision workTherefore,constructing a behavior trajectory fitting model for online public opinion subjects is a leading link in targeted control and spatial repair of public opinion,and has significant theoretical significance[Method/process] From an irrational perspective,the risk behavior of online public opinion subjects is analyzed from three dimensions:irrational internal driving behavior,irrational operational behavior,and irrational interactive behaviorThe risk behavior feature profile of online public opinion subjects is implemented from three levels:extreme emotional characteristics,sensitive information processing characteristics,and abnormal interactive behavior characteristicsBased on the extracted risk behavior features of user nodes in the user interaction space,the surface fitting of the risk behavior trajectory is achieved[Result/conclusion] The risk behavior trajectory of online public opinion subjects has been effectively fittedThe experimental results show that the risk behavior trajectory of online public opinion subjects is relatively stable during the development and dissipation periods of online public opinion,with significant fluctuations in the correlation between extreme emotional characteristics and time seriesThe overall trajectory development presents a stable risk trend in the later stageSubsequent research still needs to combine quantitative analysis of information quality and semantic relationships based on risk behavior to complete the logical loop of public opinion guidance work
  • Xu Xukan, Li Yi, Li Yiming,
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] With the rapid development of social media and the explosive increase of multimodal data,information authenticity detection faces huge challenges,and there is an urgent need to establish an effective credibility assessment model[Method/process] Based on SIP theory,this paper proposes a multimodal feature fusion method driven by fusion propagation paths to build a social media event credibility assessment modelThis method uses the graph neural network to extract the structural features of the communication network,extracts text and image features through the BERT and ResNet50 models respectively,uses the attention mechanism to achieve the fusion of multimodal features,and finally establishes a system that comprehensively considers the communication path,Weibo content and publishing a quantitative assessment framework for credibility of author characteristics[Result/conclusion] The experimental results show that compared with the single feature evaluation method,the multifeature fusion model proposed in this article has significantly improved the accuracy,recall and other indicators of credibility evaluation
  • Zhang Zhihao, Liu Sihang, Ma Tianji, An Jiajun, He Xijun
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] To facilitate precise matching of massive technology supply and demand,and to promote the conversion of innovative achievements into real productive forces[Method/process] This paper proposes a personalized recommendation method for patent transactions based on enterprise profiling,driven by Large Language ModelBased on the historical transaction information of patents by enterprises,the Large Language Model is utilized to mine and analyze the finegrained demand preference characteristics of enterprises in both the short and long termBy integrating demand preference characteristics with basic enterprise information,an enterprise profile is constructedA patent transaction recommendation method based on twostage matching is introduced,leveraging the chainofthought of Large Language Model to match the enterprise profile with patent information in the database through staged multidimensional semantic information matching,thereby achieving personalized patent transaction recommendations and providing justifications for the recommendations[Result/conclusion] Experimental results indicate that the precision,recall,and F1 score of the proposed method are 0835,0839,and 0836,respectively,significantly outperforming traditional baseline modelsUtilizing the Large Language Model’s extensive world knowledge and expertlevel reasoning capabilities can accurately construct enterprise profiles,ensuring the interpretability of recommendation results while maintaining recommendation performance[Limitations] The construction of enterprise profiles primarily relies on historical transaction information,and considering the low frequency of patent transactions,the cold start recommendation problem remains to be further addressed
  • Yang Mengqing, Zhao Yuxiang
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Digital inclusion of older adults is widely concerned by researchers but studies usually focus on their everyday information practicesUncertainty caused by the disruptions to their daily routines will bring enormous challenges when older adults encounter a life transitionThe issue of digital inclusion in such context needs to be addressed urgently[Method/process] This study is a review focusing on English literatures on the digital inclusion of older adults in the context of life transitionsWe explored the concepts of digital inclusion and life transition and provided a review of the core contexts and research methodsThen this article proposed study prospects from four aspects:focusing on innergroup heterogeneity of older adults,exploring the potential of older adults for digital inclusion,rooting in local reality of China,and promoting the integration and innovation of research methods[Results/conclusion] This article advocates exploring the issue of digital inclusion of older adults in real life transition contexts and keeping a close eye on older adults’ information behaviors and practices from the dynamic perspectives of life sceneThus,it could provide theoretical reference for bridging the digital divide of older adults
  • Guo Hongmei, Chen Yubei, Chen Xingyu, Pan Jiali, Wu Xuan, Li Guangjian
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    [Purpose/significance] Export control is a comprehensive system composed of multiple institutions and departments,laws and regulations,mechanisms and control toolsMultiple control information sources of main countries and regions are interrelatedThe article comprehensively sorts out information from multiple sources in order to provide data support for the potential impact on China’s scientific and technological development and industrial security[Method/process] The article conducts a comprehensive survey and analysis of export control mechanisms across major countries and regions,systematically compiling multidimensional information while elucidating key characteristics and interrelationships of control information resources,and practical requirements for implementing export control information systemsAnd take the USexport control list system as an example to explore the inspiration for the construction of China’s export control system[Result/conclusion] On the basis of sorting out various types and multisources of export control information,we studied the intelligence value and application services of export control resource systemThe information resources could assist to understand the evolution and changing patterns of export control policies,discover control relationships and measures between different countries and regions,grasp the technology gap and enhance the ability to warn and predict trends in key technology controlIn addition,mature export control systems abroad provide valuable experience for China to build its own similar system,which is worthy of indepth study and referenceWe will further track various export control information,improve the export control information database continuously and carry out monographic analysis of some controlled items to assist in identifying key technical points