A Novel CTR Prediction Based Model Using xDeepFM Network
Peisong Wang, Minbo Sun, Zizheng Wang, Yihang Zhou · 2021 IEEE International Conference on Computer Science, Electronic Information Engineering and Intelligent Control Technology (CEI) · 2021
Nowadays, increasing people choose to buy items online and Taobao is the largest online shopping platform. The order and the content of ads on Taobao affects the platform's revenue and the consumers' experience. Therefore, CTR(click through rate) prediction is a useful tool for enterprises to get the customers' preferences With the development of technology, there are many machine learning algorithms proposed to predict CTR, such as generalized linear model, factorization machines and deep neural network. In our paper, we utilize the xDeepFM model, which is the combination of Compressed interfere network (CIN)and Deep neural network (DNN). In experimental periods, the AUC is the metric to evaluate the prediction model's performance. xDeepFM model outperforms SVM model and neural network model by 0.017 and 0.009 respectively, which means our xDeepFM model is effective for CTR forecast.