PoP: Propagation Optimization with Popularity, Revisiting Interested Area Coverage for Mobile Advertisement

摘要

Mobile advertisement optimization has been widely studied in recent years. In this work, we focus on the propagation optimization when the content popularity is concerned. Moreover, we further investigate the issue where the number of popular advertisements is limited. These properties make the conventional optimization schemes unapplicable, because the popularity functions need to be re-evaluated with the varied popularity and decreased quantity during distribution. In dealing with aforementioned challenges, we first make statistical analysis on the popularity of the contents according to real trace data. According to our preliminary observations, we made priority evaluation for different items. We re-evaluate the distribution effects when priority is concerned. Furthermore, we improved our user selection algorithm incorporating the popularity. Evaluation results have shown the effectiveness of our approach.