基于位置的广告调度建模与求解

A New Model of Location Based Advertisement Schedule

针对车联网或未来无人驾驶场景下的多媒体广告推送问题,需要在现有的移动广告推荐系统中,加入对时间窗口、匹配路线、受众心理等参数的考虑和优化。提出了一个广告轮播模型,将一系列的广告推荐整体考虑。根据车联网环境下基于位置的广告的特点,抽象出合适的数学模型,将基于位置的广告的调度问题转化为最优化问题,并通过最优规划方法建立了广告轮播策略。实验表明该模型和策略可以使得广告与用户的时空参数最优匹配,保障广告的个性化及多样性,兼顾了广告投放者的短期利益与广告平台的长期生态效益,以及受众的心理接受度。

In order to solve the problem of multimedia advertising in the car networking or future unmanned driving scenarios,it is necessary to consider and optimize parameters such as time window,matching route,and audience psychology in the existing mobile advertisement recommendation system.An advertising rotation model was proposed to consider a whole series of advertisement recommendations as a whole.According to the characteristics of location-based advertising in the Internet of Vehicles environment,an appropriate mathematical model was abstracted,the scheduling problem of location-based advertising was transformed into the optimization problem,and the advertising carousel strategy was established through the optimal planning method.Experiments show that the model and strategy can make the optimal match between advertising and user's space-time parameters,protect the individuality and diversity of advertising,consider the short-term benefits of advertising distributors and the long-term ecological benefits of advertising platforms,and also the audience's psychological acceptance.