[1]聂鑫,孙晓燕,陈杨,等.基于改进Wide&Deep交互特征提取的移动APP转化率预估[J].郑州大学学报(工学版),2020,41(06):26-32.[doi:10.13705/j.issn.1671-6833.2020.06.005]
　SUN Xiaoyan,NIE Xin,BAO Lin,et al.The Mobile APP Conversion Rate Prediction Based on Improved Wide&Deep of Interactive Feature Extraction[J].Journal of Zhengzhou University (Engineering Science),2020,41(06):26-32.[doi:10.13705/j.issn.1671-6833.2020.06.005]
点击复制

基于改进Wide&Deep交互特征提取的移动APP转化率预估()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
41
期数:
2020年06期
页码:
26-32
栏目:
出版日期:
2020-12-31

文章信息/Info

Title:
The Mobile APP Conversion Rate Prediction Based on Improved Wide&Deep of Interactive Feature Extraction
作者:
聂鑫; 孙晓燕; 陈杨; 暴琳
中国矿业大学信息与控制工程学院,江苏徐州221008, 中国矿业大学信息与控制工程学院,江苏徐州221008, 中国矿业大学信息与控制工程学院,江苏徐州221008, 中国矿业大学信息与控制工程学院,江苏徐州221008

Author(s):
SUN Xiaoyan;  NIE Xin;  BAO Lin;  CHEN Yang
School of Information and Control Engineering, China University of Mining and Technology, Xuzhou 221008, China
关键词:
Keywords:
conversion rate prediction;  feature interaction;  field-aware factorized machine;  Wide&; Deep;  mobile APP
DOI:
10.13705/j.issn.1671-6833.2020.06.005
文献标志码:
A
摘要:
移动APP广告转化率预估已成为当前影响广告投放效率、广告排序和收益的关键因素，由于该问题特征维度高而稀疏、特征间高度交互等，使得转化率精准预估面临极大挑战。该文提出一种融合场感知分解机（Field-aware Factorized Machine, FFM）和深度卷积神经网络的改进Wide&Deep模型，以有效获取高维度稀疏特征的低阶和高阶交互关系，从而实现特征自动高效组合，提高移动APP广告转化率预估精度。在给出算法框架基础之上，针对稀疏数据的嵌入，提出了基于宽度模块FFM挖掘低阶特征交互关系的特征组合算法；然后，根据FFM所提取隐特征向量，进一步给出了基于深度模块多层卷积神经网络提取高阶交互关系的特征提取策略；最后，将宽度和深度模块分别获取的特征组合用于转化率预估。所提算法在腾讯移动APP广告转化率预估中的应用表明了该方法在提高预估精度上的有效性。
Abstract:
The conversion rate prediction (CRP) of mobile APP advertising was challenging due to the high-dimension, sparsity, and high interactions. Motivated by this an improved Wide&Deep model was proposed by fusing field-aware factorized machine (FFM) and deep convolutional neural network (DCNN) to effectively and automatically obtain the lower-order and higher-order interactions of high-dimensional sparse features, so as to realize the automatic and efficient combination of features and improve the accuracy of CRP. The framework of the proposed algorithm was first delivered, and a feature combination algorithm based on the width module FFM to extract the interactive relations of lower-order features was presented for the embedded sparse data. The extraction of the higher-order interactive features based on a DCNN was further given by fusing the latent features obtained by the FFM. Finally, the interactive feature combinations obtained by width and depth modules were integrated for the CRP. The application of the proposed algorithm in predicting the conversion rate of Tencent′s mobile APP advertisements demonstrated the effectiveness of the method.

参考文献/References:

[1] YAO Y,ZHAO W X,WANG Y,et al.Version-aware rating prediction for mobile app recommendation[J].ACM transactions on information systems,2017,35(4):38.

[2] CHAPELLE O,MANAVOGLU E,ROSALES R.Simple and scalable response prediction for display advertising[J].ACM transactions on intelligent systems and technology,2015,5(4):1-34.
[3] CHENG H T,ISPIR M,ANIL R,et al.Wide & deep learning for recommender systems[C]//Proceedings of the 1st Workshop on Deep Learning for Recommender Systems-DLRS 2016.New York:ACM,2016:7-10.
[4] RENDLE S. Factorization machines[C]//2020 IEEE International Conference on Data Mining. New York: IEEE, 2011:995-1000.
[5] JUAN Y,ZHUANG Y,CHIN W S,et al.Field-aware factorization machines for CTR prediction[C]//Proceedings of the 10th ACM Conference on Recommender Systems.New York:ACM,2016:43-50.
[6] QU Y R, CAI H, REN K, et al. Product-based neural networks for user response prediction[C]//IEEE International Conference on Data Mining. New York: IEEE, 2016: 1149-1154.
[7] WANG R X,FU B,FU G,et al.Deep & cross network for ad click predictions[C]//Proceedings of the ADKDD′17. New York:ACM,2017:1-7.
[8] CHANG Y W, HSIEH C J, CHANG K W, et al. Training and testing low-degree polynomial data mappings via linear SVM[J]. Journal of machine learning research, 2010,11(11):1471-1490.
[9] KASAP Ö Y,TUNGA M A.A polynomial modeling based algorithm in top-N recommendation[J].Expert systems with applications,2017,79:313-321.
[10] 孙晓燕，朱利霞，陈杨. 基于可能性条件偏好网络的交互式遗传算法及其应用[J]. 郑州大学学报(工学版), 2017,38(6):1-5.
[11] SUN X Y,GONG D W,JIN Y C,et al.A new surrogate-assisted interactive genetic algorithm with weighted semisupervised learning[J].IEEE transactions on cybernetics,2013,43(2):685-698.
[12] ZHU J,SHAN Y,MAO J C,et al.Deep embedding forest:forest-based serving with deep embedding features[C]//Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.New York:ACM,2017:1703-1711.
[13] LECUN Y,BENGIO Y,HINTON G.Deep learning[J].Nature,2015,521(7553):436-444.
[14] BOWERS S, BOWERS S, BOWERS S, et al. Practical lessons from predicting clicks on ads at facebook: eighth international workshop on data mining for online advertising[C]//Eighth International Workshop on Data Mining for Online Advertising. New York:ACM, 2014:1-9.
[15] RENDLE S.Factorization machines with libFM[J].ACM transactions on intelligent systems and technology,2012,3(3):1-22.
[16] BLONDEL M, FUJINO A, UEDA N, et al. Higher-order factorization machines[C]//International Conference on Neural Information Processing Systems. Massachusettc:MIT, 2016:3359-3367.
[17] XIAO J,YE H,HE X N,et al.Attentional factorization machines:learning the weight of feature interactions via attention networks[C]//Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence.Melboerne:IJCAI,2017:3119-3125.
[18] HINTON G E,DENG L,YU D,et al.Deep neural networks for acoustic modeling in speech recognition:the shared views of four research groups[J].IEEE signal processing magazine,2012,29(6):82-97.
[19] SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[EB/OL].(2015-04-10)[2020-02-04].https://arxiv.org/abs/1409.1556.
[20] NIU Y,LU Z,WEN J,et al.Multi-modal multi-scale deep learning for large-scale image annotation[J].IEEE transactions on image processing,2019,28(4):1720-1731.
[21] ZHANG W N,DU T M,WANG J.Deep learning over multi-field categorical data[M]//Lecture Notes in Computer Science.Cham:Springer International Publishing,2016:45-57.

更新日期/Last Update: 2021-02-10