Wasserstein Distance Regularized Sequence Representation for Text Matching in Asymmetrical Domains

TL;DR

This paper introduces WD-Match, a novel text matching method for asymmetrical domains that uses Wasserstein distance regularization to produce more indistinguishable feature vectors, improving matching accuracy.

Contribution

The paper proposes a Wasserstein distance-based regularizer for text matching in asymmetrical domains, enhancing existing methods by encouraging feature vector indistinguishability across domains.

Findings

WD-Match outperforms baseline methods on four benchmarks.

Regularization improves feature vector indistinguishability.

Method is compatible with multiple existing matching models.

Abstract

One approach to matching texts from asymmetrical domains is projecting the input sequences into a common semantic space as feature vectors upon which the matching function can be readily defined and learned. In real-world matching practices, it is often observed that with the training goes on, the feature vectors projected from different domains tend to be indistinguishable. The phenomenon, however, is often overlooked in existing matching models. As a result, the feature vectors are constructed without any regularization, which inevitably increases the difficulty of learning the downstream matching functions. In this paper, we propose a novel match method tailored for text matching in asymmetrical domains, called WD-Match. In WD-Match, a Wasserstein distance-based regularizer is defined to regularize the features vectors projected from different domains. As a result, the method enforces the feature projection function to generate vectors such that those correspond to different domains cannot be easily discriminated. The training process of WD-Match amounts to a game that minimizes the matching loss regularized by the Wasserstein distance. WD-Match can be used to improve different text matching methods, by using the method as its underlying matching model. Four popular text matching methods have been exploited in the paper. Experimental results based on four publicly available benchmarks showed that WD-Match consistently outperformed the underlying methods and the baselines.