| 黄颖,袁佳,叶冬梅,张慧.基于动态网络探测的技术生命周期识别研究[J].数字图书馆论坛,2025,21(4):61~71 |
| 基于动态网络探测的技术生命周期识别研究 |
| Technology Lifecycle Identification Based on Dynamic Network Detection |
| 投稿时间：2024-12-21 |
| DOI：10.3772/j.issn.1673-2286.2025.04.007 |
| 中文关键词: 技术生命周期；动态网络；网络分析；引文网络；共类网络 |
| 英文关键词: Technology Lifecycle; Dynamic Network; Network Analysis; Citation Network; Co-Classification Network |
| 基金项目:本研究得到国家自然科学基金面上项目“多源数据融合视角下技术会聚的形成机制与预测评估研究”（编号：72374162）资助。 |
| 作者 | 单位 | | 黄颖 | 武汉大学信息管理学院；武汉大学科教管理与评价中心 | | 袁佳 | 武汉大学信息管理学院；武汉大学科教管理与评价中心 | | 叶冬梅 | 武汉大学信息管理学院；武汉大学科教管理与评价中心 | | 张慧 | 武汉大学信息管理学院；武汉大学科教管理与评价中心 |
|
| 摘要点击次数: 22 |
| 全文下载次数: 51 |
| 中文摘要: |
| 识别技术的生命周期阶段对于预测技术发展趋势、制定科技政策和企业战略具有重要意义。提出一种基于动态网络探测的技术生命周期识别方法，该方法融合动态引文网络和共类网络，通过网络指标的量化曲线和分段拟合模型，划分技术生命周期阶段，全面评估技术发展态势。以薄膜晶体管液晶显示（TFTLCD）和纳米生物传感器（NBS）为例，验证了所提方法的可行性和有效性，并总结了基于动态网络的生命周期划分规则，为技术生命周期识别提供了一定参考。研究表明，TFT-LCD和NBS技术的网络划分结果与实际生命周期阶段虽存在一定的时序偏差，但仍具有高度一致性。此外，动态网络探测方法存在2~4年的滞后窗口，这与专利公开的时滞特性相关。相较于传统的S曲线法和多指标法，动态网络探测法展现出更高的鲁棒性和精确性，能避免对数据质量的过度依赖，更有效地揭示技术发展阶段，为技术生命周期的识别与预测提供了新的理论框架和实践参考。 |
| 英文摘要: |
| Accurately identifying the lifecycle stages of a technology is essential for predicting technological trends, formulating science and technology policies, and guiding corporate strategies. This study proposes a novel technology lifecycle identification method based on dynamic network detection, which integrates dynamic citation networks and co-classification networks. By quantifying network indicators and applying segmented fitting models, the method divides technology lifecycle stages and comprehensively assesses technological development trends. Using thin-film transistor liquid-crystal display (TFTLCD) and nano-biosensor (NBS) technologies as cases, this study verifies the feasibility and effectiveness of the proposed method and summarizes lifecycle classification rules based on dynamic networks, providing a theoretical and practical reference for technology lifecycle identification. The results indicate that the network-based classification of TFT-LCD and NBS technologies aligns closely with their actual lifecycle stages, albeit with a temporal offset. Furthermore, the dynamic network detection method exhibits a lag window of 2–4 years, consistent with the time delay characteristic of patent disclosures. Compared with traditional S-curve and multi-indicator methods, the dynamic network detection method demonstrates higher robustness and accuracy, reducing dependence on data quality while more precisely reflecting technological development stages. This study offers a novel theoretical framework and practical guidance for identifying and predicting technology lifecycles. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
