基于域矩阵因子分解机的点击通过率预估增强网络
Enhanced network for CTR prediction based on field-matrixed factorization machines
陈乔松 1黄泽锰 1胡静 1王进 1邓欣1
作者信息
- 1. 重庆邮电大学 计算机科学与技术学院,重庆 400065;重庆邮电大学 数据工程与可视计算重点实验室,重庆 400065
- 折叠
摘要
有效的特征交互,对于工业推荐系统中点击通过率(click-through-rate,CTR)预估的准确性起着至关重要的作用.以往并行结构的CTR预估模型通过将独立的浅层模型和深层模型并行连接,以此来学习特征的低阶交互和高阶交互.但是,这些模型存在浅层模型准确性低、未考虑特征交互时的多语义问题、参数过多、深层模型过度泛化等问题.基于上述问题,提出了一种基于域矩阵因子分解机的点击通过率预估增强网络,通过引入域矩阵优化浅层模型中的交互,提高运算效率,并在深层模型的DNN层与层之间增加了桥接模块,在每层高阶交互后增强对原始特征的记忆能力,将浅层模型和深层模型的结果相加并归一化得到预测值.该模型在Criteo、KKBox、Frappe和MovieLens数据集上进行了大量实验,展现了优秀的预测能力.
Abstract
Effective feature interaction plays a vital role in the accuracy of click-through-rate(CTR)estimation in industri-al recommendation systems.Previous CTR prediction models with a parallel structure learn low-order and high-order interac-tions of features by connecting independent shallow models and deep models in parallel.However,these models have prob-lems such as low accuracy of shallow models,failure to consider the multi-semantic problem of feature interaction,exces-sive parameters,and over-generalization of deep models.Based on the above problems,this paper proposes an enhanced network for CTR prediction based on field-matrixed factorization machines.It introduces domain matrix to optimize the inter-action in shallow models,improves the efficiency of computation,and adds a bridge module between the DNN layers of deep models to enhance the memory ability of original features after each high-order interaction.The results of shallow and deep models are added and normalized to obtain the predicted value.The model has undergone extensive experiments on Criteo,KKBox,Frappe,and MovieLens datasets,demonstrating excellent predictive capabilities.
关键词
点击通过率/域矩阵因子分解机/桥接模块/特征交互Key words
click-through rate/field-matrixed factorization machine/bridging module/feature interaction引用本文复制引用
基金项目
国家重点研发项目(2022YFE0101000)
出版年
2024
国家科技期刊平台
NSTL
万方数据