This paper presents CLEAR, a retrieval model that seeks to complement classical lexical exact-match models such as BM25 with semantic matching signals from a neural embedding matching model. CLEAR explicitly trains the neural embedding to encode language structures and semantics that lexical retrieval fails to capture with a novel residual-based embedding learning method. Empirical evaluations demonstrate the advantages of CLEAR over state-of-the-art retrieval models, and that it can substantially improve the end-to-end accuracy and efficiency of reranking pipelines.
@article{arxiv.2004.13969,
title = {Complementing Lexical Retrieval with Semantic Residual Embedding},
author = {Luyu Gao and Zhuyun Dai and Tongfei Chen and Zhen Fan and Benjamin Van Durme and Jamie Callan},
journal= {arXiv preprint arXiv:2004.13969},
year = {2021}
}