Click-Conversion Multi-Task Model with Position Bias Mitigation for Sponsored Search in eCommerce

TL;DR

This paper introduces two novel models, PACC and PACC-PE, that effectively mitigate position bias in eCommerce sponsored search, improving CTR and CVR predictions and resulting in fairer, more accurate ranking systems.

Contribution

The paper proposes two new position-bias-free models for CTR and CVR prediction, utilizing probability decomposition and neural network embeddings, to enhance ranking fairness and effectiveness.

Findings

Models outperform existing methods in ranking effectiveness.

Significantly reduce position bias in CTR and CVR predictions.

Improve fairness and accuracy in eCommerce search rankings.

Abstract

Position bias, the phenomenon whereby users tend to focus on higher-ranked items of the search result list regardless of the actual relevance to queries, is prevailing in many ranking systems. Position bias in training data biases the ranking model, leading to increasingly unfair item rankings, click-through-rate (CTR), and conversion rate (CVR) predictions. To jointly mitigate position bias in both item CTR and CVR prediction, we propose two position-bias-free CTR and CVR prediction models: Position-Aware Click-Conversion (PACC) and PACC via Position Embedding (PACC-PE). PACC is built upon probability decomposition and models position information as a probability. PACC-PE utilizes neural networks to model product-specific position information as embedding. Experiments on the E-commerce sponsored product search dataset show that our proposed models have better ranking effectiveness and can greatly alleviate position bias in both CTR and CVR prediction.