Research of Users' VOD Viewing Behaviour Based on ARMA Model

摘要

High-definition TV and interactive services bring a new viewing experience for the audience, and services such as video-on-demand and TV of yesterday have began to occupy a certain part of the viewing market gradually. Based on actual viewing records, this paper compares the differences between users' viewing time on live TV, video-on demand service and TV of yesterday service and employs the ARMA model to analyze users' VOD viewing behaviour. The results show that the overall users' VOD viewing behaviour presents a certain periodicity and users are more likely to enjoy VOD service on weekends than workdays. And the time series of viewing length on VOD per day is short-time related, so previous week's data can be used to predict future. The results of this paper can help TV network operators make strategy for business development.