Distribution Distance Regularized Sequence Representation for Text Matching in Asymmetrical Domains

Projecting the input text pair into a common semantic space where the matching function can be readily learned is an essential step for asymmetrical text matching. In the practice, it is often observed that the feature vectors from asymmetrical texts show a tendency to be gradually undistinguishable in the semantic space as the model is trained. However, the phenomenon is overlooked in existing studies. As a result, the feature vectors are constructed without any regularization, which inevitably hinders the learning of the downstream matching functions. In this paper, we first exploit the phenomenon and propose DDR-Match, a novel matching framework tailored for asymmetrical text matching. Specifically, in DDR-Match, a distribution distance-based regularizer is devised to accelerate the fusion of sequence representations corresponding to different domains in the semantic space. Then, we provide three instances of DDR-Match and make a comparison among them. DDR-Match is compatible with existing text matching methods by incorporating them as the underlying matching model. Four popular text matching methods are exploited in the paper. Extensive experimental results based on five publicly available benchmarks showed that DDR-Match consistently outperformed its underlying methods.

I. Introduction

Asymmetrical text matching, which computes the relevance score between textual documents from different domains, is a widely researched area in natural language processing (NLP) and information retrieval (IR). The wide interest of asymmetrical text matching covers a broad spectrum of real-world applications. For example, in natural language inference (NLI), text matching is used to determine whether a hypothesis is an entailment, contradiction, or neutral given a premise [2]. In question answering (QA), text matching is used to determine whether an answer can answer the given question [3], [4]. In IR, text matching is widely used to measure the relevance of a document to a query.