论文部分内容阅读
本文以TEI@I方法论为指导,提出了一个季播电视综艺节目收视率预测的研究框架.季播电视综艺节目是中国电视行业近三年发展的新兴趋势,收视率预测研究对于其排编优化和广告资源科学定价具有重要的指导意义.本文在传统数据的基础上,加入了百度指数和新浪微指数,通过建立线性回归模型发现如下规律:首期收视率对后期收视率具有锚定作用;平均收视率呈现逐年下降趋势;每年冬季和每周周五易出现收视高峰;百度指数和新浪微指数与收视率存在显著正相关.除了线性回归模型外,本文还建立了RBF神经网络、支持向量回归模型,并进行了模型集成预测,实证结果表明:加入百度指数和新浪微指数能够提高预测精度,集成模型比单一模型更能有效地预测节目收视率的走势.
This article, based on the methodology of TEI @ I, proposes a research framework for the quarterly TV variety show ratings forecasting.Child TV variety shows are emerging trends in the past three years in China’s television industry. And scientific pricing of advertising resources have important guiding significance.Based on the traditional data, we add the Baidu Index and Sina Micro Index, found the following rules by establishing a linear regression model: the first phase of the ratings on the late ratings have an anchoring role; The average viewing rate showed a declining trend year by year.Over the annual winter and weekly festivals there appeared the peak of viewing ratings.The Baidu Index and Sina Micro Index had a significant positive correlation with the viewing rate.In addition to the linear regression model, RBF neural network and support vector regression model , And carried out the model integration prediction. The empirical results show that adding the Baidu Index and Sina Micro Index can improve the prediction accuracy, and the integrated model can forecast the program ratings more effectively than the single model.