IPTV Advertising Scheduling Methods Considering Advertising Repetition Effects
研究了IPTV广告在保证合同下如何调度，提出用观看节目信息分类用户，并设计双层优化模型和REHCG算法，以最小化未交付和代表性不足，同时最大化预期买家数。
Internet protocol television (IPTV) advertising has the characteristics of both traditional TV and online advertising and mainly sells impressions through signed guaranteed contracts with advertisers. In this study, we applied the reach and frequency (R&F) contract to the IPTV sales model. Through the R&F contract, advertisers can specify the expected number of unique individuals reached and the frequency cap of the advertisements seen by each user. Considering the lack of user demographic information, we propose using viewing program information instead as the basis for user classification in targeted advertising. Advertising allocation models for two-layer optimization are established. In the upper layer optimization, the advertising allocation proportion is optimized for all user types to minimize the publisher's under-delivery and nonrepresentativeness of advertising delivery among user types. In the lower layer optimization, the impact of repeated advertising on user purchase probability is considered. The specific allocation of advertisements is optimized for each user type to maximize the expected number of buyers without violating the optimization results of the upper layer optimization. A repeated exposure-based hierarchical column generation (REHCG) algorithm is used to address the IPTV advertising scheduling problem. Numerical experiments based on actual data from the IPTV industry show that the REHCG algorithm obtains high-quality results for IPTV advertising scheduling.