SparTerm: Learning Term-based Sparse Representation for Fast Text Retrieval
This work addresses the need for efficient and interpretable text retrieval in industrial applications by enhancing semantic matching while maintaining the advantages of sparse representations.
The paper tackled the problem of improving term-based sparse representations for text retrieval by transferring knowledge from pre-trained language models, resulting in a novel framework called SparTerm that significantly outperformed traditional sparse methods and achieved state-of-the-art ranking performance on the MSMARCO dataset.
Term-based sparse representations dominate the first-stage text retrieval in industrial applications, due to its advantage in efficiency, interpretability, and exact term matching. In this paper, we study the problem of transferring the deep knowledge of the pre-trained language model (PLM) to Term-based Sparse representations, aiming to improve the representation capacity of bag-of-words(BoW) method for semantic-level matching, while still keeping its advantages. Specifically, we propose a novel framework SparTerm to directly learn sparse text representations in the full vocabulary space. The proposed SparTerm comprises an importance predictor to predict the importance for each term in the vocabulary, and a gating controller to control the term activation. These two modules cooperatively ensure the sparsity and flexibility of the final text representation, which unifies the term-weighting and expansion in the same framework. Evaluated on MSMARCO dataset, SparTerm significantly outperforms traditional sparse methods and achieves state of the art ranking performance among all the PLM-based sparse models.