Cross Domain Regularization for Neural Ranking Models Using Adversarial Learning

TL;DR

This paper introduces an adversarial learning approach to improve neural ranking models' ability to generalize across different domains by discouraging domain-specific feature learning.

Contribution

It proposes using an adversarial discriminator as a cross-domain regularizer to enhance neural ranking models' domain generalization capabilities.

Findings

Improved performance on unseen domains with up to 30% gain in precision@1.

Adversarial training effectively discourages domain-specific feature learning.

Consistent improvements across multiple experimental setups.

Abstract

Unlike traditional learning to rank models that depend on hand-crafted features, neural representation learning models learn higher level features for the ranking task by training on large datasets. Their ability to learn new features directly from the data, however, may come at a price. Without any special supervision, these models learn relationships that may hold only in the domain from which the training data is sampled, and generalize poorly to domains not observed during training. We study the effectiveness of adversarial learning as a cross domain regularizer in the context of the ranking task. We use an adversarial discriminator and train our neural ranking model on a small set of domains. The discriminator provides a negative feedback signal to discourage the model from learning domain specific representations. Our experiments show consistently better performance on held out domains in the presence of the adversarial discriminator---sometimes up to 30% on precision@1.