数据驱动的评价范式实证研究  

Empirical Research on Data-Driven Evaluation Paradigm

作　　者：李洁[1] 徐建刚[1] 黄晨[1] Li Jie

机构地区：[1]浙江大学图书馆,浙江杭州310027

出　　处：《情报理论与实践》2021年第6期55-60,共6页Information Studies:Theory & Application

基　　金：中国工程院基金项目“全球智库评价方法研究”的成果,项目编号:2018-XZ-39。

摘　　要：大数据和人工智能的发展,促进了数据分析方法的演变,改变了传统的认知方式,为形成数据驱动的科学研究新范式提供了新的契机。“数据密集型科学”的第四范式从方法论层面引发了评价范式的变革,拓展了评价的对象、指标、技术方法和建模计算,使评价对象趋向于做出更全方位立体的评价。针对各类型评价中数据采集的烦琐性和缺乏数据规范性的问题,文章设计实现了一种基于数据驱动的评价范式,以全球智库评价项目为例,提出了该评价范式的技术框架,并对能够模块化实现的数据归一化处理、权重分析、基于缺失值的补缺和线性加权计算等做了介绍。最后以实际项目验证了该数据驱动评价范式的可行性。The advancement of big data and artificial intelligence has promoted the development of data analysis methods,created new cognitive methods,and provided strong support for the formation of a new paradigm of data-driven scientific research.The fourth paradigm of“data-intensive science”triggers a change in the evaluation paradigm from the methodological level,expanding the evaluation objects,indicators,technical methods and modeling calculations,so that the evaluation subjects tend to make a more comprehensive evaluation.Aiming at solving the problems of the cumbersome data collection and lack of data standardization in various types of evaluation,an evaluation paradigm based on data-driven is designed and implemented.This paper takes the“Global Think Tank Evaluation”project as an example to propose the technical framework of the evaluation paradigm.Data normalization processing,weight analysis,filling missing values and linear weighting calculation are also introduced.Finally,the feasibility of the data-driven evaluation paradigm was verified with the actual projects.

关 键 词：数据驱动 评价范式 通用性 智库评价 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论] TP18[自动化与计算机技术—计算机科学与技术]