基于时间序列模型的收视率研究

收视率是指在某个时段收看某个节目的目标观众人数占总目标人群的比重,以百分比表示。收视率的高低能够反映出节目的观众喜爱程度以及节目所具有的社会影响力。本文采用2006年中国电视收视年鉴中的全国收视数据,以工作日及节假日收视率为研究对象,利用SAS统计软件做时间序列分析,并进行预测。结果显示,节假日收视率数据能够较好的拟合ARIMA(1,2,(1,4))模型,工作日收视率数据则利用ARIMA(1,2,(1,4))模型和AR(1,4)模型拟合结果都较好,但是相对而言ARIMA(1,2,(1,4))拟合效果最好。

Study on the Ratings Based on Time Sequence Model

Ratings refers to the proportion of target audience in total population,when they watch a show in a certain period of time,expressed by percentage. The level of ratings can refect the degree of audience avorite and the social influence of the program. Based on the rating number of the whole country,in“2006 chinese television viewers yearbook”,this paper analysed the ratings of working days and holidays,using the SAS system software. The results show that time sequence model ARIMA(1,2,(1,4)) fits the holi-day ratings data very well,ARIMA(1,2,(1,4))model and AR(1,4)model fit the working days ratings data both very well. By contrast,the ARIMA(1,2,(1,4)) model is best.