节点文献

基于神经网络的视频网站自制节目播放量的预测建模

Modeling on Playback Volume Prediction of Self-produced Programs of Chinese Video Websites

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 李鉴桥; 关虎; 张树武;

【Author】 LI Jian-qiao;GUAN Hu;ZHANG Shu-wu;Information Engineering School,Communication University of China;Digital Content Center,Institute of Automation,Chinese Academy of Sciences;

【机构】 中国传媒大学信息工程学院; 中国科学院自动化研究所数字内容中心;

【摘要】 自2005年起,由最初的低成本到如今的专业规模化,我国视频网站自制节目以其形式多元、内容丰富成为众多网民的新宠。本文选取五大视频平台,共101部数据进行分析,提出了一套影响播放量的指标体系,并且在该指标体系基础之上构建了一套基于BP神经网络的播放量预测模型,同时针对神经网络波动性的特点,对预测模型进行必要的改进。该指标体系和预测模型的研究能为节目上映前的投资、盈利提出更全面可靠的参考方案,对我国视频产业的研究与发展有所裨益;而且,以该播放量预测模型为基础,利用科学、合理的评估体系和方法,可以对视频版权相关产品或服务进行定价,有助于简化并规范版权交易的询价议价过程,科学合理地确定版权作品的创意附加值,对版权价值评估及版权交易的进行具有重要指导意义。

【Abstract】 Since 2005,from the initial low-cost to today’s professional scale,self-produced programs of Chinese video websites become new favorite of many net citizens depend on its form of diversity and substantial content. This paper selects 5 major video websites,analyses 101 data,a index system of an impact at the play times was proposed,and build a model on BP neural network based on the index system. At the same time,according the volatility of BP,the forecasting model is improved. The research of the index system and the model can provide a more comprehensive and reliable reference scheme for the investment and profit before the release of the program,which is beneficial to the research and development of China’s video industry. Moreover,based on the forecasting model,the use of scientific and reasonable evaluation system and methods can be used to price video related products or services to help simplify and standardize the bargaining process of copyright transactions,scientifically and rationally. The creative added value of copyright works is of great significance to the evaluation of copyright and the conduct of copyright transactions.

【基金】 国家科技支撑计划课题(2015BAH49F01)
  • 【文献出处】 中国传媒大学学报(自然科学版) ,Journal of Communication University of China(Science and Technology) , 编辑部邮箱 ,2017年06期
  • 【分类号】G222;TP183
  • 【被引频次】6
  • 【下载频次】408
节点文献中: 

本文链接的文献网络图示:

本文的引文网络