情报科学 ›› 2023, Vol. 41 ›› Issue (6): 94-102.

• 业务研究 • 上一篇    下一篇

基于DySAT模型的时序动态技术融合预测
——以中国5G通信领域为例

  

  • 出版日期:2023-06-01 发布日期:2023-07-27

  • Online:2023-06-01 Published:2023-07-27

摘要:

【目的/意义】随着中国5G通信技术的发展进入快车道,将有更多的技术与其发生技术融合,提前预测技术
融合趋势对国家政策布局和企业决策具有重要意义。【方法/过程】基于DySAT模型,引入时间维度信息建立时序动
态技术融合预测框架,对中国5G通信领域进行技术融合预测。【结果/结论】基于DySAT模型的时序动态技术融合
预测框架可以提高技术融合预测性能,在长期与短期中均取得较好的预测效果,对5G通信领域的实证分析验证了
该方法的可行性。预测结果显示,5G通信领域中的技术融合主要涉及智慧交通、智慧医疗以及智慧教育三大方向。
【创新/局限】从不同时间段的网络中提取结构特征并且进行特征聚合,将先前时刻的网络作为网络序列进行技术
融合预测,弥补了过往研究中未考虑网络演变过程的不足。

Abstract:

【Purpose/significance】 As the development of 5G communication technology of China enters the fast lane, there will be
more technology convergence within it. It is of great significance to predict the trend of technology convergence in advance for policy layout and enterprise decision making.【Method/process】 Based on the DySAT model, this paper constructs temporal dynamic technol⁃ogy convergence prediction framework by considering the information of time dimension and predict technology convergence in 5G communication field of China.【Result/conclusion】 The temporal dynamic technology convergence prediction framework based on the DySAT model can improve the performance of technology convergence prediction in both long-term and short-term. The prediction re⁃sults show that the technology convergence in 5G communication field mainly reflect in intelligent transportation, intelligent health⁃care, and intelligent education.【Innovation/limitation】 Extracting and aggregating structural features from the networks in different time periods, using previous networks for technical convergence prediction can overcome the deficiency of not considering the evolu⁃tion process of the network in the previous research.