基于网络搜索的票房预测模型——来自中国电影市场的证据  

Forecasting box office performance based on online search:Evidence from Chinese movie industry

作　　者：王炼[1,2] 贾建民[2,3] 

机构地区：[1]西南石油大学经济管理学院,成都610500 [2]西南交通大学经济管理学院,成都610031 [3]香港中文大学商学院

出　　处：《系统工程理论与实践》2014年第12期3079-3090,共12页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(71090402);西南石油大学青年教师"过学术关"计划

摘　　要：建立了一个将网络搜索与电影票房联系起来的模型,考察了影片上映前后消费者网络搜索的动态变化及其在电影票房预测中的作用,文章就中国电影产业对该模型进行了实证检验.结果显示,影片网络搜索量以及增长趋势都能预测该影片的首映周票房.对于后续周票房,动态联立方程模型的估计结果表明,网络搜索既是电影票房的先行信号,也是电影票房的后向结果.样本外分析表明,整合了网络搜索的模型能够提高预测的准确性.We develop a modeling framework that links online search and box office revenues in order to examine the dynamic patterns of consumer online search and its role in predicting the market performance of motion pictures both before and after theatrical releases. We empirically test the model in the context of the Chinese motion picture industry. Our findings provide evidence that both the volume and the growing pattern of online search for a movie have predictive power on its opening-week market performance. Results of a dynamic simultaneous-equations model further reveal that online search is both a precursor and an outcome of sales in the later run of a movie＇s lifetime. Out-of-sample analysis shows that incorporating online search helps make more accurate forecasts.

关 键 词：网络搜索 票房 预测 动态联立方程 

分 类 号：F272.1[经济管理—企业管理]