TY - JOUR
T1 - Webpage Depth Viewability Prediction Using Deep Sequential Neural Networks
AU - Wang, Chong
AU - Zhao, Shuai
AU - Kalra, Achir
AU - Borcea, Cristian
AU - Chen, Yi
N1 - Funding Information:
This work is partially supported by NSF under grants No. CAREER IIS-1322406, CNS 1409523, and DGE 1565478, by a Google Research Award, and by an endowment from the Leir Charitable Foundations. Any opinions, findings, and conclusions expressed in this material are those of the authors and do not necessarily reflect the views of the funding agencies. Chong Wang and Shuai Zhao contributed equally to this work.
Publisher Copyright:
© 2019 IEEE.
PY - 2019/3/1
Y1 - 2019/3/1
N2 - Display advertising is the most important revenue source for publishers in the online publishing industry. The ad pricing standards are shifting to a new model in which ads are paid only if they are viewed. Consequently, an important problem for publishers is to predict the probability that an ad at a given page depth will be shown on a user's screen for a certain dwell time. This paper proposes deep learning models based on Long Short-Term Memory (LSTM) to predict the viewability of any page depth for any given dwell time. The main novelty of our best model consists in the combination of bi-directional LSTM networks, encoder-decoder structure, and residual connections. The experimental results over a dataset collected from a large online publisher demonstrate that the proposed LSTM-based sequential neural networks outperform the comparison methods in terms of prediction performance.
AB - Display advertising is the most important revenue source for publishers in the online publishing industry. The ad pricing standards are shifting to a new model in which ads are paid only if they are viewed. Consequently, an important problem for publishers is to predict the probability that an ad at a given page depth will be shown on a user's screen for a certain dwell time. This paper proposes deep learning models based on Long Short-Term Memory (LSTM) to predict the viewability of any page depth for any given dwell time. The main novelty of our best model consists in the combination of bi-directional LSTM networks, encoder-decoder structure, and residual connections. The experimental results over a dataset collected from a large online publisher demonstrate that the proposed LSTM-based sequential neural networks outperform the comparison methods in terms of prediction performance.
KW - Computational advertising
KW - recurrent neural networks
KW - sequential prediction
KW - user behavior
KW - viewability prediction
UR - http://www.scopus.com/inward/record.url?scp=85047624393&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85047624393&partnerID=8YFLogxK
U2 - 10.1109/TKDE.2018.2839599
DO - 10.1109/TKDE.2018.2839599
M3 - Article
AN - SCOPUS:85047624393
SN - 1041-4347
VL - 31
SP - 601
EP - 614
JO - IEEE Transactions on Knowledge and Data Engineering
JF - IEEE Transactions on Knowledge and Data Engineering
IS - 3
M1 - 8362690
ER -