A CTR prediction model based on attention mechanism and logarithmic transformation

Lu Meng; Tian-Wei Shi; Guang-Ming Chang; Jiao-Feng Qiang; Wen-Hua Cui

doi:10.1117/12.2641350

4 August 2022 A CTR prediction model based on attention mechanism and logarithmic transformation

Lu Meng, Tian-Wei Shi, Guang-Ming Chang, Jiao-Feng Qiang, Wen-Hua Cui

Proceedings Volume 12306, Second International Conference on Digital Signal and Computer Communications (DSCC 2022); 123061M (2022) https://doi.org/10.1117/12.2641350
Event: Second International Conference on Digital Signal and Computer Communications (DSCC 2022), 2022, Changchun, China

Abstract

To address the problems of inadequate feature interaction and lack of targeting in feature combination in the clickthrough rate prediction model. We propose a click-through prediction model called SELFM. It based on attention mechanism and logarithmic transformation structure. The model first incorporates the attention mechanism in the feature embedding stage to distinguish the importance of different features and avoids the effects of invalid features. Then the field-aware factorization machine is used to learn low-order feature interactions. The logarithmic transformation structure is used to convert the power of each feature in the feature combination into the coefficients to be learned and combined with the hidden layer for higher-order nonlinear feature interactions. The final output layer is processed with the Sigmoid function to get the click-through rate prediction results. The experimental results show that the AUC and Logloss of this paper's model are better than the existing click-through rate prediction models, which effectively improves the prediction accuracy and enhances the ability of the recommendation system to process data.

Citation Download Citation

Lu Meng, Tian-Wei Shi, Guang-Ming Chang, Jiao-Feng Qiang, and Wen-Hua Cui "A CTR prediction model based on attention mechanism and logarithmic transformation", Proc. SPIE 12306, Second International Conference on Digital Signal and Computer Communications (DSCC 2022), 123061M (4 August 2022); https://doi.org/10.1117/12.2641350

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available