In this paper, we focus on examining the effects of Ad-context on the click-Through rate (CTR) for the online advertising. Many researches have shown that ad-context congruity is a key factor to CTR, but the features of ad-context are rarely introduced in CTR prediction algorithms. By leveraging data from various sources, and using text mining and sentiment analysis techniques, our proposed approach extracts three types of features (i.e. users, advertisements and ad-context features) to predict CTR. User features describe "Who" is browsing the webpage, advertisement features depict "How" an ad serving is, and ad-context features include "What" the product is, "Where" an ad displays, as well as the "Mood" of the context. Experiment results show that our proposed approach outperforms the benchmark models by introducing ad-context features. Novel ad-context features we proposed make good contributions to the prediction of CTR. The research highlights the power of ad-context in online advertising CTR prediction. Moreover it gives an insight into supporting effective advertising and improving users' satisfaction.
CITATION STYLE
Sun, C., Zhang, M., & Zuo, M. (2018). Does ad-context matter on the effectiveness of online advertising? In Proceedings of the International Conference on Software Engineering and Knowledge Engineering, SEKE (Vol. 2018-July, pp. 439–444). Knowledge Systems Institute Graduate School. https://doi.org/10.18293/SEKE2018-144
Mendeley helps you to discover research relevant for your work.