基于GCA-RFR模型的数字内容资源价值评估方法研究  

Research on Value Evaluation Method of the Digital Content Resources Based on GCA-RFR Model

作　　者：赵艳 王文举[2] 倪渊[3,4] ZHAO Yan;WANG Wen-ju;NI Yuan(School of Economics,Beijing Technology and Business University,Beijing 100048,China;School of Economics,Beijing Wuzi University,Beijing 101149,China;School of Economics and Management,Beijing Information Science & Technology University,Beijing 100192,China;Laboratory of Bid Date Decision Making for Green Development,Beijing 100192,China)

机构地区：[1]北京工商大学经济学院,北京100048 [2]北京物资学院经济学院,北京101149 [3]北京信息科技大学经济管理学院,北京100192 [4]绿色发展大数据决策北京市重点实验室,北京100192

出　　处：《统计与信息论坛》2022年第2期12-22,共11页Journal of Statistics and Information

基　　金：国家重点研发计划青年科学家项目“文化产品产权价值评估与确权标识应用技术研究”(2021YFF0900200);北京市组织部优秀人才(青年拔尖项目);北京工商大学青年教师科研启动基金资助项目。

摘　　要：随着数字内容资源数量和重要性的不断增加,逐渐暴露出很多问题,如数字内容产品质量参差不齐、数字内容资源交易混乱、价值衡量缺乏规范性等,因此探索合理的价值评估方法对规范数字内容资源市场交易具有重要意义。首先,根据“价值链理论”梳理影响数字内容资源价值的成本因素、版权因素、市场因素、服务因素等,并从中选出13个二级指标构建数字内容资源价值评估指标体系。然后,在该指标体系基础上提出一种基于GCA-RFR模型的智能化价值预测方法,依次经过指标验证、样本筛选、模型训练等步骤对电影资源数据进行实证分析,并将该模型的预测效果与GCA-BP模型进行对比,确定了该预测模型的优越性。研究发现,基于GCA-RFR模型的方法系统化保证了其在数字内容资源价值评估方面体现的独特优势:强客观性、优泛化性、调节参数少、预测精度高等。该方法的探索为丰富数字内容资源价值评估理论体系拓展了方向。As the number and importance of the digital content resources continue to increase,many problems are gradually exposed,such as uneven quality of digital content products,chaotic transactions of digital content resources,lack of normative value measurement,etc.Therefore,exploring a reasonable value evaluation method is of great significance for regulating the market transactions of the digital content resources.First,according to the“value chain theory”,it sorts out the cost factors,copyright factors,market factors,service factors,etc.That affect the value of the digital content resources,and selects 13 secondary indicators from them to construct a digital content resource value evaluation index system.Then,on the basis of the indicator system,an intelligent value forecasting method based on the GCA-RFR model is proposed.And the empirical analysis of the movie resource data is carried out through the steps of indicator verification,sample screening,and model training.Finally,the forecasting effect is compared with the GCA-BP model,and the superiority of forecasting model is determined.The study found that the GCA-RFR model systematically guarantees its unique advantages,such as strong objectivity,excellent generalization,few adjustment parameters,and high prediction accuracy,in the digital content resources value evaluation.The innovation of this study is the application of GCA-RFR model.First,a framework for evaluating the value of digital content resources is constructed with the index system.Under the framework,generalized grey relational analysis can be used to preliminarily screen and verify the rationality of each index setting.Then,the idea of optimizing the samples is adopted,and the sample sets with high correlation are selected from the obtained large number of samples,so as to constitute the initial data of the training model.It can ensure that the trained evaluation model reduces the systematic error caused by the data as much as possible.Finally,choosing the more mature random forest model as the

关 键 词：数字内容资源 价值评估 指标体系 灰色关联分析 熵值法 随机森林回归模型 BP神经网络 

分 类 号：F064.1[经济管理—政治经济学]