基于高阶网络的三元技术融合过程和预测研究

Process Analysis and Prediction of Triple Technology Convergence Based on Higher-order Network

[目的/意义]多技术融合已成为当前科技创新的重要特征,开展多元技术融合过程和预测研究,识别技术融合的趋势和规律,有助于支撑国家战略决策、企业研发策略制定。[方法/过程]提出一种基于高阶网络的技术融合预测方法,通过三元技术融合过程分析和高阶链路预测,预测特定领域的多元技术融合。首先,以专利分类号为节点,采用单纯复形方法构建专利技术高阶网络;其次,结合边的形成和加强,分析三节点组的形态和三角形闭包事件的过程;最后,构建训练集和测试集,从12个模型中获得效果最佳的相似性指标,开展三元技术融合预测。[结果/结论]以人工智能领域进行实证研究,发现边的加强对于三角形闭包事件具有重要意义,成功预测了F16M、G06F、G06N等三节点组发生三元技术融合,验证了方法的有效性。

[Purpose/significance] Multi-technology convergence has become an important feature of current science and technology innovation.This study conducts research on the process and prediction of multi-technology convergence to identify the patterns and trends of technology convergence,which helps to support national strategic decision-making and enterprise R&D strategy formulation.[Method/process] A higher-order network-based technology convergence prediction method is proposed to predict domain-specific multi-technology convergence through triple technology convergence process analysis and higher-order link prediction.Firstly,a patent technology higher-order network is constructed using simplicial complexes method with patent classification numbers as nodes.Secondly,the pattern of three-node groups and the process of triangle simplicial closure events are analyzed in conjunction with edge formation and edge strengthening.Finally,training and testing sets are constructed to obtain the similarity metrics with the best results from the 12 models to carry out triple technology convergence prediction.[Result/conclusion] An empirical study was conducted in the field of artificial intelligence,and it was found that edge strengthening is important for triangle simplicial closure events,and the three-node groups of F16M,G06F and G06N were successfully predicted to undergo triangle technology convergence,which verified the effectiveness of the method.