 |
|
|
|
|
|
| Intelligence Theories and Methods |
|
|
503 |
Influence of Interdisciplinary Collaboration of Technical Teams on Breakthrough Innovation Hot! |
|
 |
Ding Lerong, Shi Jing, Wu Keye, Sun Jianjun |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.001 |
|
|
This study investigates the impact of interdisciplinary collaboration within technical teams on breakthrough innovation. Utilizing biomedicine patent data from the PATSTAT global patent database, we assess the interdisciplinary nature of technical teams across three dimensions: variety, balance, and disparity. The degree (measured by the D index) and the quantity of breakthrough innovation are employed to qualify the achievements of these teams. Using multiple linear regression analysis, we analyze the influence of interdisciplinary collaboration within technical teams on breakthrough innovation. The results show that, in the field of biomedicine, interdisciplinary technical teams significantly impact breakthrough innovation. The higher the variety within these teams, the more likely they are to produce breakthrough innovations of high caliber and the more the breakthrough innovation achievements of the team conform to the characteristics of “small but beautiful.” Conversely, the greater the disparity within the technical team, the more likely they are to produce breakthrough innovations with low impact. Moreover, the higher the balance of the technical team, the more favorable the outcomes. We also find that the technical teams in the field of biomedicine are mainly small-scale teams with under ten people, among which young, non-transnational small-scale teams are more likely than others to produce high breakthrough innovation results. |
|
|
2024 Vol. 43 (5): 503-515
[Abstract]
(
131
)
HTML
(185 KB)
PDF
(1351 KB)
(
169
) |
|
|
|
516 |
Research on Interdisciplinary Paper Recommendation Based on Academic Network Hot! |
|
|
Du Jin, Xiong Huixiang, Xiang Yinghong |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.002 |
|
|
To meet the needs of information science researchers for interdisciplinary papers, this study develops an interdisciplinary paper recommendation model based on academic network. First, according to the keyword coupling network and author citation network characteristics, the correlation between the author and the paper is extracted, facilitating paper recommendation based on keyword coupling. Second, this model utilizes the author citation coupling network, encompassing cross-disciplinary citation relationships, co-citation relationships, and the subject attribute of the paper. This information is used to mine interdisciplinarity aspects of authors and papers, calculating interdisciplinary similarity for subject-based recommendations. Finally, the study integrates recommendations based on keyword coupling and subject similarity to achieve a hybrid recommendation for interdisciplinary papers. The model is validated using data from Chinese Social Sciences Citation Index (CSSCI) database. Empirical results demonstrate that the recommended papers exhibit interdisciplinary characteristics. Compared with keyword coupling-based recommendations, combining interdisciplinary characteristics improves author recommendation success rates, average accuracy rates, and average recall rates. |
|
|
2024 Vol. 43 (5): 516-527
[Abstract]
(
146
)
HTML
(239 KB)
PDF
(1232 KB)
(
101
) |
|
|
|
528 |
Research on the Rapid Response Mechanism of Intelligence in a Complex Information Environment Hot! |
|
|
Su Cheng, Zhao Runbo, Huang Yanning, Zhao Xiaoyuan |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.003 |
|
|
With the rapid development of science and technology and the intensification of great power games, intelligence has become an important tool for seizing national development opportunities. This study examines the realization of intelligence support in national strategic decision-making in this new environment of intelligence production and evolving requirements of decision-making. Through literature research, case study analysis, and other methods, this study summarizes the relevant theories and practical experience based on the connotation, characteristics, and decision-making support scenarios of the intelligence rapid response mechanism for in-depth investigation. This study advances an innovative “3+5+9 Rapid Response Mechanism of Intelligence” applicable to the complex information environment. The mechanism takes “early discovery,” “early prediction,” “early reporting,” “early decision-making,” and “early response” as the five core objectives, and “linkage coordination,” “information monitoring,” “information resources,” “intelligence analysis,” “emergency response,” “technical support,” “expert collaboration,” “organizational safeguards,” and “tracking and evaluation supervision” as the nine support subsystems. This intelligence working mechanism is applicable to three important decision-making support scenarios: regular monitoring, responding to concerns, and sudden emergencies, to provide solid theoretical guidance for future practice. |
|
|
2024 Vol. 43 (5): 528-537
[Abstract]
(
64
)
HTML
(90 KB)
PDF
(1007 KB)
(
302
) |
|
|
|
538 |
Construction of a Knowledge Representation Model in Public Health Emergencies for Information Disclosure Hot! |
|
|
Xiang Yafan, Liu Dongsu, Ma Xubu, Qin Chunxiu, Shi Ying |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.004 |
|
|
In recent years, there have been frequent public health emergencies that have posed great challenges to government emergency management. The timely disclosure of event information by the government helps eliminate public panic, which is crucial for epidemic prevention and control as well as socio-economic development. However, information on public health emergencies is scattered across a variety of locations in a fragmented, discontinuous, and incomplete manner. The description and organization of such a massive, diverse, and changing body of information that is not well-integrated is thus key to government emergency management. Therefore, this study is oriented toward information disclosure and integrates the knowledge graph and event knowledge graph, constructs a knowledge representation model of public health emergencies, and represents the core concepts and relationships within it. It also provides a method to organize information on public health emergencies that can both represent the spatial and temporal evolution of an epidemic and display epidemic information. The results show that this model has good vertical properties and displays a relatively rich conceptual relationship and attribute features, and most classes can be filled with examples. The knowledge representation model proposed in this study provides new ideas for the construction of an emergency knowledge base, thereby expanding the methodological system of emergency information organization. Further, it also helps in disclosing and releasing the value of stock information, satisfying the public’s information needs, and improving the effectiveness of emergency management. |
|
|
2024 Vol. 43 (5): 538-552
[Abstract]
(
74
)
HTML
(152 KB)
PDF
(2986 KB)
(
76
) |
|
|
|
553 |
Identifying Moderation Effects via Meta-Analysis of Big Data: Basic Model and Empirical Testing Hot! |
|
|
Lin Weijie, Zhou Wenjie, Wei Zhipeng, Yang Kehu |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.005 |
|
|
Moderation effect testing, as an important method for identifying causal relationships in empirical research, helps uncover underlying relationships between independent and dependent variables. However, this method suffers from issues such as the inability to obtain true effect values and low external validity. Owing to the limitations imposed by the inherent flaws of primary studies, the evidence-based field urgently needs new models for identifying moderation effects. In this study, we adopt the concept of big data evidence and use a recursive method to systematically arrange and combine control variables, thereby simulating the “exhaustion” of all possible original research designs. We conduct regression analyses and record all effect values for all possible variable relationships, and then use meta-analysis to comprehensively merge all original effect sizes to obtain true effect values and enhance the external validity of moderation effect results. Finally, taking research on information poverty as an example, this study demonstrates in detail the entire process of identifying moderation effects from a big data evidence perspective. The main contribution of this paper is the enhancement of the meta-analysis framework within the realm of big data evidence-based approach. This involves distilling authentic effect sizes from an extensive compilation of original research findings, thereby augmenting the external validity of moderating effects and enhancing the dependability of causal inference. |
|
|
2024 Vol. 43 (5): 553-562
[Abstract]
(
84
)
HTML
(129 KB)
PDF
(1241 KB)
(
88
) |
|
|
| Intelligence Technology and Application |
|
|
563 |
Identification of Potential Emerging Technologies by Fusing SVM-LDA and Weighted Similarity: Taking the Field of Artificial Intelligence as an Example Hot! |
|
|
Ran Congjing, Tian Wenfang |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.006 |
|
|
In the context of a new round of technological revolution and accelerated industrial transformation, accurately identifying emerging technologies with disruptive potential in the constantly emerging technological ocean is of great significance for the nation, enterprise participants, and relevant commercial investment institutions. It is therefore important to grasp the development trends and directions of technological innovation, reasonably allocate scientific and technological resources, and carry out advance scientific and technological strategic planning and technological layout. This article proposes an emerging technology topic recognition model based on knowledge-enhanced SVM-LDA. First, a classification standard for basic technology was developed based on the prior knowledge of the expert group; second, the technology category classification criteria were input into the SVM-LDA model as a priori knowledge to obtain the technology topic clustering results; third, a weighted similarity calculation based on category subject terms was performed to identify potential emerging key technologies; and fourth, an empirical study was conducted using the field of artificial intelligence as an example. Finally, 24 potential emerging technologies were obtained, mainly distributed across six major categories: special robot technology, monitoring and early warning technology, video and image processing technology, voice recognition technology, automated planning and decision-making techniques, and natural language processing technology. |
|
|
2024 Vol. 43 (5): 563-574
[Abstract]
(
71
)
HTML
(131 KB)
PDF
(2511 KB)
(
136
) |
|
|
|
575 |
Core Patent Identification Method Integrating Attribute Indicators and Citation Relationships Hot! |
|
|
Guo Jianming, Wang Jingyi, Yuan Run |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.007 |
|
|
Based on information theory, this paper proposes a core patent identification method integrating attribute index and citation relationships for balancing individual characteristics and the network integrity of patents. First, it analyzes the necessity and feasibility of patent information analysis based on a comprehensive attribute index and citation relationship from the perspective of information theory and constructs core patent identification models based on comprehensive methods. Second, the patent index system is constructed to calculate the patent attribute value, and PageRank and HITs (hyperlink-induced topic search) algorithms are used to measure the importance, authority, and hub of patents before and after the relationship between comprehensive attribute value and direct citation, co-citation and coupling, and identify core patents. Finally, we attempt to compare the effects of the methods before and after synthesis using the method based on the robustness and vulnerability of complex networks. The empirical results indicate that (1) the identification model combining the two methods increases the amount of information for patent analysis, considers the advantages of patent index analysis and network analysis, and realizes the complementary advantages of both methods. (2) Different citation relationships reflect the difference in patent values; the identification results of the three relationships are concentrated and discrete, and a few core patents simultaneously have high importance, authority, and hub. (3) The evaluation method based on the robustness and vulnerability of complex networks is useful for solving the problem of evaluating results in patent information analysis. |
|
|
2024 Vol. 43 (5): 575-587
[Abstract]
(
53
)
HTML
(137 KB)
PDF
(2804 KB)
(
80
) |
|
|
|
588 |
Automatic Generative Information Science Term Extraction and Multidimensional Linked Knowledge Mining Hot! |
|
|
Hu Haotian, Deng Sanhong, Kong Ling, Yan Xiaohui, Yang Wenxia, Wang Dongbo, Shen Si |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.008 |
|
|
Information science terminology conveys the basic knowledge and core concepts of information science discipline. It is thus of great significance to sort out and analyze information science terms from the basic concepts to promote the development of the discipline and assist downstream knowledge mining tasks. With the rapidly growing amount of scientific and technological literature, automatic term extraction has replaced manual screening, but existing methods rely heavily on large-scale labeled datasets, making it difficult to migrate to low-resource scenarios. This study designs a Generative Term eXtraction for Information Science (GTX-IS) method, which transforms the traditional extraction task based on sequence labeling into a sequence-to-sequence generative task. Combined with few-shot learning strategies and supervised fine-tuning, it improves the ability to generate text for specific tasks and can more accurately extract information science terms in low-resource scenarios. For the extraction results, this study further develops term discovery and multi-dimensional knowledge mining in the field of information science, and comprehensively uses full-text informetric and scientometric methods to conduct statistical analysis and knowledge mining on the frequency of occurrence, life cycle, and co-occurrence information of terms from the dimensions of the term itself, the relationship between terms, and time information. Using the social network analysis method, combined with the characteristics of the time dimension, this study improves the dynamic profile of journals, facilitating the exploration of the research hotspots, evolution process, and future development trends of information science. The proposed method surpasses all 13 baseline generative and extractive models, showing a strong few-shot learning ability, and provides a new idea for domain information extraction. |
|
|
2024 Vol. 43 (5): 588-600
[Abstract]
(
75
)
HTML
(152 KB)
PDF
(5124 KB)
(
118
) |
|
|
|
601 |
Topic Classification of Ancient Texts Based on SWPF2vec and DJ-TextRCNN Hot! |
|
|
Wu Shuai, Yang Xiuzhang, He Lin, Gong Zuoquan |
|
|
DOI: 10.3772/j.issn.1000-0135.2024.05.009 |
|
|
The method for classifying topics in ancient book texts, mainly based on cataloging and rule matching, encounters challenges such as low efficiency, heavy reliance on expert knowledge, a single classification basis, and difficulties in automating the classification process. In addressing these issues, this study attempts to classify themes that meet the researchers’ needs based on the content and characteristics of ancient texts, and promote the transformation of digital humanities research paradigms. First, referring to the analysis method of characters in the ancient book Analytical Dictionary of Characters (Shuowen Jiezi) of the Eastern Han Dynasty, a new four-dimensional feature dataset of “pronunciation (speaking) - original text (text) - structure (pattern) - glyph (font)” is constructed based on the corpus dataset of ancient books. Second, a four-dimensional feature vector extraction model (speaking, word, pattern, and font to vector; SWPF2vec) is designed and combined with a pre-trained model to achieve fine-grained feature representation of ancient texts. Once again, the ancient text topic classification model (dianji - recurrent convolutional neural networks for text classification; DJ-TextRCNN) is constructed by fusing convolutional neural networks, recurrent neural networks, and multi-head attention mechanism. Finally, multidimensional, deep-level, and fine-grained semantic mining of ancient texts is achieved by integrating four-dimensional semantic features. DJ-TextRCNN exhibits the best accuracy in topic classification under different dimensional features, achieving an accuracy of 76.23% under the four-dimensional feature of “shuo, wen, jie, zi,” preliminarily achieving accurate topic classification of ancient book texts. |
|
|
2024 Vol. 43 (5): 601-615
[Abstract]
(
46
)
HTML
(203 KB)
PDF
(2050 KB)
(
69
) |
|
|
|
