文章摘要
张 越 曹 悦.基于复杂网络分析的技术预测弱信号识别研究[J].中国科技资源导刊,2025,(3):48~58,69
基于复杂网络分析的技术预测弱信号识别研究
Research on Detection of Weak Signals for Technical Forecasting Using Complex Network Analysis
投稿时间:2024-10-22  
DOI:
中文关键词: 弱信号识别;技术预测;复杂网络;连通分量
英文关键词: weak signal detection, technology forecasting, complex network, connected components
基金项目:中国科学技术信息研究所创新研究基金青年项目“非线性动力学视角下的技术监测弱信号识别研究”(QN2024-04);国家重点研发计划项目“颠覆性技术感知响应平台研发与应用示范之地平线扫描系统”(2019YFA0707202)
作者单位
张 越 曹 悦 (中国科学技术信息研究所,北京 100038) 
摘要点击次数: 15
全文下载次数: 45
中文摘要:
      为克服传统弱信号识别方法在捕捉新兴技术趋势方面存在的局限性,提出一种基于图连通特征的弱信号识别优化模型。利用自然语言处理技术,提取和筛选文献关键词,并进行词频和增长趋势分析。通过构建关键词共现网络,利用复杂网络分析中连通分量和社区检测算法,自动识别关键词之间的紧密关系并将其分组,增强弱信号的识别效果。对未来深海深空深地建筑领域的技术预测进行实证分析。结果表明,所提出的模型在识别弱信号的准确性和实用性方面显著提升,为科技领域的前瞻性研究提供了强有力工具。未来将结合更多异构数据增强对弱信号特征的全面性揭示,提升在其他新兴领域弱信号识别的准确性。
英文摘要:
      Traditional methods of weak signal detection face limitations in capturing emerging technical trends. This study proposes an optimized approach to weak signal detection based on graph connectivity features. Natural language processing techniques were employed to extract and filter keywords from academic literature, followed by analyses of word frequency and growth trends. By constructing a keyword co-occurrence network and applying complex network analysis methods, such as connected component and community detection algorithms, the close relationships among keywords were automatically identified and clustered, thus enhancing the effectiveness of weak signal detection. Empirical results from the technical forecasting in the future deep- sea , deep-space and deep-earth architecture fields demonstrate that the proposed model significantly enhances the accuracy and practical utility of weak signal detection, providing a valuable tool for foresight studies in the science and technology domain. Future work will focus on incorporating more heterogeneous data to further enhance the comprehensive identification of weak signals, thereby improving the accuracy of weak signal detection across other emerging fields.
查看全文   查看/发表评论  下载PDF阅读器
关闭

分享按钮