摘要
技术成熟度评价在科技项目管理工作中得到越来越深入的应用,然而,当前评价工作依赖于大量评价信息的人工评审和分析,缺乏评价数据的标准化处理及存储,导致评价效率较低,且难以开展相关机器学习辅助评价模型的开发。基于项目技术成熟度评价原始数据,开展数据内容共性分析,提出了一种基于Vision Transformer框架的数据处理方法,并设计了一种新的数据结构,实现技术成熟度评价数据的标准化及标准化数据库构建,最终通过Fleiss-Kappa分数验证了该数据库的真实性,该方法能够优化数据存储模式,提高数据质量,形成了机器学习模型的数据基础。
Technology maturity evaluation has been applied more and more deeply in the management of science and technology projects.However,the current evaluation work relies on manual review and analysis of a large amount of evaluation information,and lacks standardized processing and storage of evaluation data.As a result,the evaluation efficiency is low,and it is difficult to develop relevant machine learning-assisted evaluation models.Based on the original data of project technology maturity evaluation,data content commonality analysis is carried out,a data processing method based on the Vision Transformer framework is proposed,and a new data structure is designed,thus,the standardization of technology maturity evaluation data and the construction of standardized database are realized.Finally,the authenticity of the database is verified by Fleiss-Kappa score.This method can optimize the data storage mode,improve the data quality,and form the data basis of the machine learning model.
作者
杨雪鹤
杨颖坤
莫冰
YANG Xuehe;YANG Yingkun;MO Bing(CEPREI,Guangzhou 511370,China;Guangdong Provincial Key Laboratory of Electronic Information Products Reliability Technology,Guangzhou 511370,China;Key Laboratory of Active Medical Devices Quality&Reliability Management and Assessment,Guangzhou 511370,China)
出处
《电子质量》
2024年第3期121-128,共8页
Electronics Quality
关键词
数据标准化
数据库构建
深度学习
技术成熟度评价
data standardization
the construction of database
deep learning
technology readiness evaluation