Scheduling web banner advertisements with conflicts and time windows: a column generation-based solution approach
摘要
In the ever-evolving landscape of online advertising, where revenue holds significant importance for Internet-based organizations, the efficient scheduling of banner advertisements has become paramount. This paper studies the banner advertisement scheduling problem, expanding its scope to include ad conflicts and display time windows, highlighting the complex dynamics of modern advertising. Conflicts arise when advertisements cannot be shown in the same time slots, often due to brand rivalry or content incongruity. Furthermore, each accepted ad must be displayed within a designated time window. The study develops a column generation-based algorithm to tackle this challenging scheduling problem. Computational experiments on large datasets show that the column generation approach is highly effective, delivering near-optimal solutions in a reasonable time frame. This efficiency is crucial in an industry where time-sensitive decisions to maximize revenue are paramount.