Special Topics

1009 Data Intelligence Empowered Innovation: An Exploration of the Innovation Assistance Framework Based on Data Intelligence Technology Hot! Lu Wei, Ma Yongqiang, Liu Jiawei, Yang Jinqing, Cheng Qikai DOI: 10.3772/j.issn.1000-0135.2023.09.001 Large language models (LLMs), such as ChatGPT (Chat Generative Pre-trained Transformer), have shown excellent performance in text generation and human-machine dialog. Under the background of LLMs, such technologies as big data and artificial intelligence have demonstrated important practical value in empowering scientific research and innovation. Although current science and technology (S&T) information resource management and knowledge services can provide relatively accurate information and routine knowledge aggregation services for scientific research and innovation, they have not yet been deeply integrated with scientific research and innovation activities. Researchers also face challenges such as insufficient information processing capabilities and limited cognitive abilities during the scientific research and innovation process. This article analyzes the new characteristics of scientific research activities in the data intelligence era. Furthermore, an innovation assistance framework based on data intelligence technologies is proposed, and its functional positioning, service mode, and key empowerment path in the entire innovation process are analyzed and discussed in depth. In the future, with the continuous maturity and progress of big data and artificial intelligence technologies, data-intelligence-technology-enabled S&T information resource management will be further embedded in the entire process of scientific research and innovation activities. Innovative assistance services based on data intelligence technologies, such as innovative auxiliary services for literature reading, experiment design, and article-writing scenarios, can provide researchers with personalized, fine-grained knowledge and scenario-based solutions to better serve scientific research and innovation activities. 2023 Vol. 42 (9): 1009-1017 [Abstract] ( 418 ) HTML (93 KB)  PDF (2000 KB)  ( 599 )

Intelligence Theories and Methods

1018 Identifying Emerging Scientific Topics by Abrupt Change of Knowledge Structure Hot! Duan Qingfeng, Chen Hong, Yan Xuxian, Liu Dongxia DOI: 10.3772/j.issn.1000-0135.2023.09.002 Understanding the structural transformation of knowledge is key to identify emerging scientific topics. Novel distribution of knowledge elements implies that emerging topics occur, and transformation of knowledge structure has become the characteristic to identify emerging topics. Following these ideas, we measured the difference of structure between two knowledge networks in sequential time using the WL subtree kernel and propose an indicator in terms of structure transformation of knowledge to suggest the extent to which a new topic emerges. In addition, we propose another indicator that measures the growth rate of a topic’s influence over time using the PageRank algorithm. Then, these two indicators together compose a two-dimensional space for identification and form an integrated solution based on the dynamic topology of the knowledge network. The empirical research in the field of information science well-validated our method, with high sensitivity and effective discrimination towards emerging topics with high value in the short term. These topological transformation-based indicators are capable of quantitatively analyzing the evolution of academic topics and offer a unique view of knowledge structure to enable scientific evolution. 2023 Vol. 42 (9): 1018-1028 [Abstract] ( 185 ) HTML (125 KB)  PDF (1886 KB)  ( 214 )

1029 Empirical Study of Knowledge Association Methods for Major Scientific Discoveries Hot! Ren Xiaoya, Zhang Zhiqiang DOI: 10.3772/j.issn.1000-0135.2023.09.003 Scientific discovery is that of “undiscovered discoveries,” and major scientific discoveries can promote the exploration and development of science. The emergence of scientific discovery is a complex phenomenon, and scientific knowledge is embedded therein. However, owing to the proliferation of information; limitations of knowledge; and artificial removal, concealment, or weakening, more undiscovered implicit relationships between scientific discoveries exist. Therefore, we present a knowledge association method of scientific discoveries that combines multiple citation relations to determine the knowledge-based associations between major scientific discoveries from a deeper granularity and a more diverse perspective. We adopt the Fields Medal, Lasker Award, and Turing Award as the empirical research objects. Quantitative and time series analyses are used to explore the knowledge dissemination regularities and characteristics of scientific discovery between various fields from the perspective of backtracking, combined with a qualitative interpretation of scientific discoveries and expert consultation. Finally, we refine the knowledge dissemination patterns of scientific discovery. Our results reveal five types of knowledge dissemination patterns between scientific discoveries: linear (simple and closed linear), bridge, radial, and multiple. Our empirical findings could deepen researchers’ understanding of the regularities of scientific discoveries and provide a reference for policy formulation in research funding and scientific evaluation. 2023 Vol. 42 (9): 1029-1039 [Abstract] ( 213 ) HTML (90 KB)  PDF (2510 KB)  ( 278 )

1040 Construction of Super-Network for Emergency Information Cooperation Hot! Zhang Xinrui, Zhang Haitao, Luan Yu, Zhang Chunlong DOI: 10.3772/j.issn.1000-0135.2023.09.004 The world is currently undergoing profound, unprecedented changes. The COVID-19 outbreak severely tested the world, making the era of digital wisdom considerably unpredictable and replete with challenges``. Developing methods for improving the efficiency of emergency management has become a key issue. This study innovatively explored the information cooperation process based on multi-agent collaboration in emergencies from the perspective of a super network and analyzed the internal logic of information cooperation. Further, this study constructed a multi-agent collaborative network based on the collaborative relationship between multiple agents, the co-evolution between text information networks, text by keywords co-occurrence relationships between formation information gene networks, and the mining and defining of distinct longitudinal dynamic mapping relationship between the three types of network. A super-network model was eventually constructed for emergency information cooperation; moreover, a visual analysis and verification were conducted on actual public health emergency cases. This study achieved an in-depth understanding of emergency information, identified the core information gene with collaborative values, mined core agents and texts, and analyzed the dynamic and evolutionary nature of the information collaboration process. Ultimately, the findings of this study promote the construction of a deep and panoramic information collaborative network relationship, provide new research ideas and methods for the subsequent exploration of practical application, enable collaborative decision-making in emergencies under the concept of national security with intelligence wisdom, provide constructive suggestions for social collaborative governance, and make emergency management digital and intelligent. 2023 Vol. 42 (9): 1040-1051 [Abstract] ( 234 ) HTML (160 KB)  PDF (5477 KB)  ( 268 )

1052 Measure of Disruptive Innovation in Science: Relative Disruptive Index (RDI) Hot! Yang Alex J., Deng Sanhong, Wang Hao DOI: 10.3772/j.issn.1000-0135.2023.09.005 Innovation serves as a propulsive impetus behind scientific advancement. Assessing the innovative merit of scientific research holds particular significance. Papers endowed with elevated levels of innovation often challenge and upend existing disciplinary paradigms within research fields. A newly introduced metric called disruptive index has emerged as a means of directly appraising the innovation encapsulated within papers. By leveraging the profound interplay of paper citations and the network of cited relationships, this index effectively surmounts the limitations associated with unidimensional evaluations. As a result, it has garnered considerable attention within the scientific metrology community. This study employed a comprehensive synthesis and expansion of the latest research pertaining to the disruptive index. It provides a profound analysis of the limitations and influencing factors associated with this index. Additionally, it proposes a novel metric called relative disruptive index (RDI), which builds upon the foundation laid by the disruptive index. By encompassing the deep quotation proportions derived from the citation network, the RDI strives to mitigate the issues stemming from the non-uniform distribution and inconsistent evaluation encountered within the realm of the disruptive index. Consequently, it engenders a more precise and objective means of gauging the innovation inherent within scientific research endeavors. Empirical results clearly demonstrate the inconsistency in parameter magnitudes across the disruptive index. Conversely, the relative disruptive index emerges as a more efficacious gauge of research innovation compared to the original disruptive index and DI5. It is confirmed that the RDI exhibits superior consistency in evaluation, yielding a lower relative subversion index for papers that build upon previous work and higher relative subversion index for papers that embody groundbreaking innovation. 2023 Vol. 42 (9): 1052-1064 [Abstract] ( 392 ) HTML (238 KB)  PDF (2498 KB)  ( 320 )

Intelligence Technology and Application

1065 Emotion Prediction of Network Public Opinions Based on the Deviation Rules Markov Model Hot! Shi Wei, Xue Guangcong, He Shaoyi DOI: 10.3772/j.issn.1000-0135.2023.09.006 Using sentiment analysis technology, the evolution of online public opinion can be predicted from the perspective of emotion transfer in short microblog comments. In this study, the short texts of related microblog comments in the early stage of the COVID-19 pandemic were taken as the research object. Based on the extended association rule Apriori Algorithm and Markov Chain, a new method called the deviation rule Markov model is proposed. This model analyzes the correlation and transfer between Internet users’ emotion classes and predicts the changing trends of Internet users’ emotional states in the early stage of the pandemic by calculating the transfer probability of different emotion classes and constructing a time-varying emotion state transfer matrix. The experimental results demonstrated that the emotional state of netizens after the pandemic outbreak was not negative but gradually changed to “positive” emotions over time. Through a comparative analysis of examples, the validity and accuracy of the affective prediction model for online public opinion proposed in this study were verified. 2023 Vol. 42 (9): 1065-1077 [Abstract] ( 207 ) HTML (162 KB)  PDF (3746 KB)  ( 179 )

1078 What Causes the Failure of Emergency Decision Making? Configuration Analysis in the Intelligence View Hot! Sun Qinying, Ma Haiqun DOI: 10.3772/j.issn.1000-0135.2023.09.007 The paradigm of emergency decision making has been upgraded from passive emergency response to active preventive preparation, which focuses on failure traceability and prevention now. Emergency decision-making failure is influenced by multiple concurrent factors. Mining the influencing factors of such failure from the perspective of intelligence is conducive not only to grasping the complex relationship between emergency intelligence and decision in theory, but also to understanding the decision-making subjects’ thinking and behavior, which has practical guiding value for improving the efficiency of emergency decisions. Based on Simon’s three-stage model of management decision making, this study adopts the qualitative comparative analysis (QCA) method to extract seven explanatory variables from aspects of intelligence collection, decision design, and program selection. This study analyzes the configuration of 44 emergency cases in China, and further explores the failure path leading to the emergency decision. The results indicate the following. From the single conditional variable analysis, government emergency decision-making failure is significantly affected by incomplete intelligence integration and insufficient cross-department coordination. From the analysis of conditional configuration, the path of government emergency decision failure can be divided into three modes: decision deviation and fault, lag avoidance, and utility failure. This study presents countermeasures to prevent the failure of emergency decision making from constructing the emergency intelligence perception mechanism of three-dimensional space, online and offline dual path triggering mechanism of technology enablement, auxiliary decision-making mechanism of multi-party cooperation, and assessment and accountability mechanism of multi-ability improvement. Furthermore, it provides a reference for the government to solve the dilemma of the failure of emergency decision making and improves the efficiency of decision making, which has important theoretical and practical significance in revealing the differentiated path of the failure of such decision making. 2023 Vol. 42 (9): 1078-1091 [Abstract] ( 173 ) HTML (214 KB)  PDF (2438 KB)  ( 275 )

1092 Characteristics of “Malicious Terms” in Privacy Policies and Their Interactive Mechanisms Based on Content Analysis Hot! Zhang Mingxin, Zhu Hou DOI: 10.3772/j.issn.1000-0135.2023.09.008 With the profit-driven nature of business organizations, Internet platforms tend to comply publicly with privacy protection laws but privately violate user rights using ambiguous and unequal terms in privacy policies. Previous research has explored the content characteristics and format design of privacy policies but failed to reveal malicious terms in these policies. To fill this gap, we selected 19 representative privacy policies from 19 Internet platforms. Fourteen categories (five main categories and one core category) were identified. Based on their predominant discourses, the results demonstrate that platforms adopt strategies such as shifting responsibility and excessively requesting personal information to obtain users’ automatic acceptance. Considering the results, this study proposes several suggestions, including enhancing user engagement in policy design, designing more interactive user interfaces, and constructing mechanisms of dynamic acceptance, to provide implications for regulating the process of designing privacy policies and enhancing personal information protection. 2023 Vol. 42 (9): 1092-1102 [Abstract] ( 147 ) HTML (116 KB)  PDF (914 KB)  ( 201 )

Intelligence Users and Behavior

1103 Multi-perspective Affective Theory Models and Their Applied Expansion in the Field of Information Behavior Hot! Zhang Min, Zhang Ke DOI: 10.3772/j.issn.1000-0135.2023.09.009 Affect is a proven key factor in understanding human information behavior. However, the research on affect is still at a relatively early stage, characterized by confusing terminology, a single research perspective, and lack of systematic analysis of emotion function, and exploration of the interaction mechanism between affect and information behavior is still relatively rudimentary. The multi-perspective theory integration needs to be strengthened. This paper presents four perspectives of affective theory according to the historical development and philosophical origin: affect feeling, affect motivation, affect evaluation, and affect information. Further, the application and development of related models in the field of information behavior were compared and analyzed. By identifying and summarizing the core ideas of four theoretical perspectives, including the analysis and deconstruction of the triggers, components, and structures, and the relationship between affective functions and cognitive behavior, an integrated conceptual framework is constructed, and further expanded research paths are proposed to provide reference for subsequent in-depth research in this field. 2023 Vol. 42 (9): 1103-1113 [Abstract] ( 230 ) HTML (142 KB)  PDF (1907 KB)  ( 289 )

Intelligence Discipline Development and Construction

1114 Construction of Core Competence of Information Science in the Transformation Period of Discipline Development Hot! Xie Li, Chen Yunwei, Lu Ying, Xia Dong DOI: 10.3772/j.issn.1000-0135.2023.09.010 There have been many academic disputes in the development of Information Science in China, reflecting the problems faced by discipline construction, such as the discipline orientation being unclear, core theories being weak, and research contents being extensive. Big data and AI together break discipline boundaries, promote cross integration, and drive changes to the scientific research paradigm, forcing information research and work to embrace the new pattern; they carefully examine and systematically summarize the core competitiveness of the discipline. From a historical perspective, we organized the key issues in the development of information science, such as the origin of discipline, development of X-metrics, and evolution of discipline territory, and explored the “inheritance” of disciplines behind phenomena in terms of discipline tasks, research objects, research content, etc. We also analyzed the “innovation” in the research paradigm of the discipline when facing a new environment and new situation by analyzing big data-driven and technological enablement. Thus, we rethought the construction of informatics discipline. The core competence of information science lies in the industry instinct of putting forward core information issues, the professional integrity of improving the efficiency of information acquisition as its own responsibility, and an open mind in embracing cross-border cross-integration of demand changes. Informationists should continue to consolidate information ideas and theories, expand information research thoughts and methods, and cultivate cross-border integration capabilities. 2023 Vol. 42 (9): 1114-1124 [Abstract] ( 152 ) HTML (113 KB)  PDF (1171 KB)  ( 280 )

Intelligence Reviews and Comments

1125 Research on Health Disinformation on the Internet Hot! Zhu Qinghua, Chen Qiong, Lu Dongmei, Wang Lei, Song Shijie, Zhao Yuxiang, Zhao Yuehua DOI: 10.3772/j.issn.1000-0135.2023.09.011 Social media platforms, such as Weibo and WeChat, have become the main channels for health information dissemination on the Internet. Health disinformation has been increasingly attracting the attention of scholars. The purpose of this paper is to promote collaborative governance against health disinformation dissemination mechanisms and contribute to the optimization of the online health information ecosystem through a literature review. The state of the current research on health disinformation is outlined from the concept and connotation perspectives, and the research conducted on the factors influencing health disinformation is reviewed, thereby identifying health disinformation intervention mechanisms. Finally, the research gaps are summarized, and future directions are suggested to preclude health disinformation. 2023 Vol. 42 (9): 1125-1138 [Abstract] ( 288 ) HTML (207 KB)  PDF (1013 KB)  ( 391 )