Intelligence Theories and Methods

251 Cooperation Tendencies of Breakthrough Innovation Inventors Hot! Yu Bowen, Liu Xiang DOI: 10.3772/j.issn.1000-0135.2024.03.001 Breakthrough technological innovation is key to promoting industrial upgrading, and R&D collaborations are an effective way to achieve technological innovation. As a natural extension of technological innovation, the analysis of the collaboration tendency among breakthrough innovation inventors helps better understand the differences and connections between breakthrough and continuous innovations, which can help promote the reasonable construction of inventor cooperation relationships and the cultivation of R&D teams. This study first defines a measure for breakthrough innovation inventors, constructing an inventor collaboration network using patent citation data, and finally analyzes and verifies the collaboration tendencies of breakthrough innovation inventors based on the null model. The study demonstrates that the inventor collaboration network displays degree correlations and assortativity, in that, breakthrough innovation inventors tend to cooperate with breakthrough innovation inventors, and the tendency of collaboration among breakthrough innovation inventors becomes increasingly obvious over time. 2024 Vol. 43 (3): 251-260 [Abstract] ( 20 ) HTML (135 KB)  PDF (1825 KB)  ( 45 )

261 A Two-Stage Approach for Technology Supply-Demand Matching Based on Text Matching and Configuration Optimization Hot! Yu Hui, Xia Wenlei, Huang Wei, Wu Yunjing, Zhang Yufan DOI: 10.3772/j.issn.1000-0135.2024.03.002 As the key link of technology transfer, technology supply and demand matching is an important measure to improve the efficiency of technology transfer, accelerate industrial upgrading, and realize the power of science and technology. Addressing the double challenge of excessive matching results and the complexity of the influencing factors in the process of technology supply and demand matching, this study proposes a model based on text matching and configuration optimization and verifies the model with real published information. The model is based on previous research results. Feedback elicited from the technical demander on the recommendation of technical achievements confirmed that as the supply and demand matching scheme of the study has a certain accuracy and feasibility, it can provide the basis for the construction of a science and technology service platform in practice. 2024 Vol. 43 (3): 261-273 [Abstract] ( 17 ) HTML (113 KB)  PDF (1831 KB)  ( 33 )

274 Health Information Persuasive Mechanism Considering Time Effect: An Empirical Research Based on Elaboration Likelihood Model Hot! Ke Qing, Ding Mengya, Cao Yaning, Li Jiawen DOI: 10.3772/j.issn.1000-0135.2024.03.003 In current health information behavior research, the pressing focus is on persuading the public to adopt good health behaviors through health information dissemination or education. Following the elaboration likelihood model (ELM), this study divided the persuasive routes into central and peripheral routes, employing a short-term longitudinal design. A diary experiment spanning 10 days was used continuous health information use data from 30 college students (total 377 datapoints). To explore the persuasive mechanism of health information on health behavior willingness, a hierarchical linear model (HLM) was established at the individual, information clue, and time levels. Results indicate that clues mainly depend on a mixed persuasion route of information quality and source credibility. Additionally, over a seven-day period, the persuasive effect gradually strengthened. Information quality exhibited a more stable persuasive effect, while the effect of source credibility is diminished over time. Moreover, the persuasion route of health information is contingent on individual characteristics and usage time. Health awareness is time-dependent and moderates the effect of source credibility and information relevance. Involvement also moderates the effect of information quality and source credibility on persuasion, with no time dependency. This study sheds light on the persuasive mechanisms in health information for health behavior change and provides suggestions for developing “people-oriented” personalized health information dissemination and education programs. 2024 Vol. 43 (3): 274-286 [Abstract] ( 9 ) HTML (188 KB)  PDF (862 KB)  ( 45 )

287 Research on Improving the Semantic Main Path Analysis Method by Leveraging the Density Peak Clustering Algorithm Hot! Chen Liang, Yu Chi, Shang Weijiao, Xu Haiyun, Lyu Shijiong, Chen Lili DOI: 10.3772/j.issn.1000-0135.2024.03.004 The semantic main path analysis (sMPA) method overcomes the shortcomings of the traditional main path analysis (MPA) method, such as a single main path and low theme consistency. However, it also leaves two defects: the position of the selected main path in the semantic space may deviate from the cluster center, and the topic discrimination of different main paths is not obvious. To address this problem, this study proposes a gradually optimized main path selection method in which topic cluster density and path traversal weight are superimposed to form a composite density, and the location of the topic cluster center is optimized by adjusting the proportion of the two elements in the composite density. When the cluster center converges, the paths located in different topic cluster centers are outputted. This method is verified by applying it to the patent citation network of lithium-ion batteries for electric vehicles and the citation network of high-impact papers in the field of materials science. The experimental results show that not only is the layout of multiple main paths generated by the new method but the possibility of selecting improper main paths is also significantly reduced. 2024 Vol. 43 (3): 287-301 [Abstract] ( 16 ) HTML (194 KB)  PDF (3963 KB)  ( 36 )

Intelligence Technology and Application

302 Weak Signal Detection and Identification with Heterogeneous Data and Semantic Construction Hot! Han Meng, Chen Yue, Wang Yuqi, Wang Kang, Cui Linwei DOI: 10.3772/j.issn.1000-0135.2024.03.005 Business competition is intensifying, and the iteration cycle of innovation and products is accelerating. Thus, identifying the weak early signals of emerging technologies is of significant strategic importance. This study explores a novel method for weak signal detection and identification using heterogeneous data and semantic construction. First, heterogeneous data are filtered through document outlines and text dissimilarities based on academic articles. Second, weak signals represented by keyword terms are extracted by improved term frequency-inverse document frequency (TF-IDF) and the combination graph method. Finally, the context of the terms is obtained through semantic construction and mining to explain the meaning of weak signals, thereby identifying the weak signals of emerging technologies. Taking the field of quantum information as an example, this study identifies the weak signals of current emerging technologies, including quantum theory of dot adsorption, private agreement, dot medicine, and quantum games. The results confirm the effectiveness of the method for weak signal detection and identification. The method is expected to be helpful for science and technology managers in research decision-making. 2024 Vol. 43 (3): 302-312 [Abstract] ( 11 ) HTML (168 KB)  PDF (1616 KB)  ( 42 )

313 Co-occurrence Hierarchical Sampling for Academic Document Representation Learning Hot! Ding Heng, Zhang Jing, Chen Jiazhuo, Cao Gaohui DOI: 10.3772/j.issn.1000-0135.2024.03.006 Effective feature representations of academic papers can be used in the classification and ranking of academic papers, thereby improving the efficiency of searches to provide users with more intelligent and effective literature recommendations and personalized services. Inspired by the study of citation proximity analysis (CPA) in information science, we utilize a self-supervised contrastive learning framework, to propose a co-citation hierarchical sampling algorithm that allows mining of potential associations among documents from structured full-text data. A self-supervised prior training task is constructed for training the citation co-occurrence hierarchical transformer (CCHT), which is an academic text representation model at the document level. The S2ORC and SPECTER training sets were used to construct triplets from co-citations of the same sentence, paragraph, and chapter to train the proposed research models, which were subsequently applied to the four major SciDocs benchmark tasks of document classification, user behavior prediction, citation prediction, and paper recommendation. Different evaluation metrics were adopted for the different tasks. Specifically, in the document classification task, the F1 metric; in the user behavior prediction and citation prediction tasks, the normalized discounted cumulative gain (nDCG) and mean average precision (MAP) metrics; and in the paper recommendation task, P@1 and nDCG^ were used for evaluation. The results demonstrate that (1) the CCHT model outperformed the other baseline models in the SciDocs benchmark test set, performing best when positive samples with fixed sampling levels were co-citations of the same sentence; (2) hierarchical citation co-occurrence based hard negative sampling may introduce noisy data during training, which degrades performance. 2024 Vol. 43 (3): 313-326 [Abstract] ( 11 ) HTML (243 KB)  PDF (2890 KB)  ( 30 )

327 A Study on Seal Recognition Method Based on Data Augmentation and Vision Transformer Hot! Zhang Zhijian, Xia Sudi, Liu Zhenghao, Wang Wenhui, Chen Shuaipu, Huo Chaoguang DOI: 10.3772/j.issn.1000-0135.2024.03.007 Seal recognition poses challenges due to difficulties in data collection, annotation, and image degradation. This study aims to alleviate data scarcity through data augmentation and improve the model's ability to recognize seals in complex scenarios by using the vision transformer (ViT) model to extract global features. First, the contextual characteristics of the seals are analyzed, implementing data augmentation strategies based on the analysis results to expand the training set. Seal images are then input into the ViT model for feature extraction and recognition. We collected and annotated 1,259 seals from 16 calligraphy and painting works, such as “Lanting Xu.” After applying 11 data augmentation modules, the training set expanded to include 127,159 seal images. Compared with the baseline model ResNet50, the F1 score improved by 12.17%. When the extended data obtained through data augmentation is removed, all models fail to converge. However, the proposed method lacks semantic reasoning ability and cannot recognize seals not present in the training set. In scenarios with limited annotated data, the combination of data augmentation techniques and the utilization of the ViT model can facilitate accurate seal image recognition. 2024 Vol. 43 (3): 327-338 [Abstract] ( 20 ) HTML (136 KB)  PDF (2434 KB)  ( 9 )

Intelligence Users and Behavior

339 Formation and Evolution Path of User's Digital Hoarding Behavior: A Dual-System Theory Perspective Hot! Jia Mingxia, Zhao Yuxiang, Zhu Qinghua, Wu Dawei DOI: 10.3772/j.issn.1000-0135.2024.03.008 Serious hoarding behavior involving the excessive acquisition of and inability to discard digital information of little or no value can disrupt individuals' learning, personal lives, and work, and may even have adverse effects on information resource management, digital governance, and the growth of data factor markets. We analyzed a mixed dataset comprising 3,568 valid comments on digital hoarding behavior collected from 21 interview transcripts and 3 social media platforms. Rooted theory was employed to elucidate the formation and evolution of individuals' digital hoarding behaviors in the context of the big data and social media environment. Our findings reveal that the development of digital hoarding behaviors is non-linear, primarily driven by the imbalance between individuals' “false sense of uncertainty avoidance” drive and “self-regulation and control” inhibitions during the process of digital accumulation. Additionally, our study identifies three types of uncertainty-seeking and control processes and identifies contextual factors that influence digital hoarding behavior. We also examine the often overlooked but widespread reflexive process of self-regulation and control within irrational information behavior to gain deeper insights into why digital hoarding persists despite coping measures. This study constructs a substantive theoretical model to explain the formation and evolution of digital hoarding behaviors based on the dual systems theory. The findings promote deeper insights into the digital hoarding phenomenon within the field of information science. This study not only provides new insights for the management decisions of information service platforms and digital information service organizations but also holds important implications for the information resource management and sustainable development of individuals, digital enterprises, organizations, and even digital society. 2024 Vol. 43 (3): 339-356 [Abstract] ( 10 ) HTML (188 KB)  PDF (1493 KB)  ( 93 )

Intelligence Discipline Development and Construction

357 Constructing the Framework of an Intelligence Theory System for Coordinating Science and Technology Development and Science and Technology Security Hot! Li Pin DOI: 10.3772/j.issn.1000-0135.2024.03.009 Coordinating technological development and security has become an important national strategy. As an important component of the technology strategy consulting system, the intelligence system should play its due role. Building an intelligence system that supports the coordinated development and security of science and technology requires the support of relevant theories. This study mainly explores the theoretical issues of constructing an intelligence system. Therefore, this study starts with the integration of technological development and security, explores the theoretical framework of the intelligence system, and deeply analyzes the three elements within its theoretical framework: the guarantee system, governance system, and operation system. 2024 Vol. 43 (3): 357-368 [Abstract] ( 8 ) HTML (105 KB)  PDF (1593 KB)  ( 29 )

369 Empowering Scientific and Technological Intelligence Education in a Complex Information Environment Hot! Su Peng, Zhang Zeyu, Lin Zijie, Wu Zhenfeng, Zeng Wen DOI: 10.3772/j.issn.1000-0135.2024.03.010 Scientific and technological intelligence is an important aspect of intelligence education. The complex information environment has introduced new challenges for scientific and technological intelligence education. Using methods such as reviewing literature of national and international intelligence education expounds the organization, presentation, and implementation of the content of empowering education, such as awareness, means, knowledge, and ethics. In view of the special requirements for the training of scientific and technological intelligence professionals, this study clarifies the “changed innovation” and “unchanged integrity” of intelligence education in a complex information environment. First, the empowerment content of science and technology intelligence education is gradually enriched, and a new version of scientific and technological intelligence science is created in the new knowledge reserve, information means, and ethical concepts. Second, the empowerment mode of scientific and technological intelligence education is increasingly adaptable, and the ability to deal with complex task scenarios of scientific and technological intelligence should be formed through lifelong learning, in-depth practice, and awareness shaping. Third, the empowerment goal of scientific and technological intelligence education is gradually clarified, creating a benign environment for the development of disciplines, transfer of talents, and transfer of value. On this basis, the author presents the content, mode, and goal of China's scientific and technological intelligence education in a complex information environment to help explore the empowerment model for scientific and technological intelligence education that suits China. 2024 Vol. 43 (3): 369-376 [Abstract] ( 13 ) HTML (96 KB)  PDF (728 KB)  ( 21 )