Neural Click Models for Recommender Systems

TL;DR

This paper introduces neural architectures inspired by click models to better understand user behavior in recommender systems, utilizing advanced neural networks like recurrent and Transformer-based models to improve performance and evaluation.

Contribution

It presents novel neural architectures for user behavior modeling in recommender systems, extending traditional click models with advanced neural network techniques.

Findings

Models outperform baseline methods on ContentWise and RL4RS datasets.

Neural models can simulate user responses for system evaluation.

Proposed architectures include recurrent, Transformer-based, adversarial, and hierarchical models.

Abstract

We develop and evaluate neural architectures to model the user behavior in recommender systems (RS) inspired by click models for Web search but going beyond standard click models. Proposed architectures include recurrent networks, Transformer-based models that alleviate the quadratic complexity of self-attention, adversarial and hierarchical architectures. Our models outperform baselines on the ContentWise and RL4RS datasets and can be used in RS simulators to model user response for RS evaluation and pretraining.