Unsupervised Document Expansion for Information Retrieval with Stochastic Text Generation
This addresses the vocabulary mismatch issue for information retrieval systems, offering an unsupervised solution that avoids the need for labeled query-document pairs, though it is incremental as it builds on existing expansion methods.
The paper tackles the vocabulary mismatch problem in information retrieval by proposing an unsupervised document expansion framework using a pre-trained language model and stochastic embedding perturbations, which significantly outperforms baselines on two benchmark datasets.
One of the challenges in information retrieval (IR) is the vocabulary mismatch problem, which happens when the terms between queries and documents are lexically different but semantically similar. While recent work has proposed to expand the queries or documents by enriching their representations with additional relevant terms to address this challenge, they usually require a large volume of query-document pairs to train an expansion model. In this paper, we propose an Unsupervised Document Expansion with Generation (UDEG) framework with a pre-trained language model, which generates diverse supplementary sentences for the original document without using labels on query-document pairs for training. For generating sentences, we further stochastically perturb their embeddings to generate more diverse sentences for document expansion. We validate our framework on two standard IR benchmark datasets. The results show that our framework significantly outperforms relevant expansion baselines for IR.