摘要
针对现有技术体系如何实现定量化的网络化建模和识别技术创新的模式和机理两个主要问题,提出数据驱动的技术创新网络构建方法 ,挖掘出技术文档的关键字向量,结合向量空间模型实现技术相似度的定量计算,生成技术创新时序网络。讨论了技术创新网络基本结构,分析了技术创新网络的模体类型与特性,并计算了模体重要性剖面以进一步确定其网络特征,得出技术创新具有强合作性且技术创新网络与生物网络以及信号传输网络属于同一个网络超家族的结论。并以技术评论数据为例验证了该方法的有效性。
To solve the quant itat ive network model ing of technology network and dist inguish the mode of technology breakthrough, a data-driven network modeling approach is proposed. The similarity between tech-nologies is computed quantitatively with vector space model and the keyword vector of technology documents which aims to model the sequential network. The basic structure and motif characteristic of technology break-through network are discussed. The significance profile is calculated to emulate the network deeply. Experimen-tal results of technology review data confirm the effectiveness of the proposed approach and show that there is a strong cooperation between innovational technologies and the network is similar with biological network and sig-nal-transduction network, which all belong to the network superfamily.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2017年第5期1072-1077,共6页
Systems Engineering and Electronics
基金
国家自然科学基金(71501182
71571185)资助课题
关键词
模体分析
技术创新
复杂网路
数据驱动
motifanalysis
technology breakthrough
complex network
data-driven
