Corpus-Steered Query Expansion with Large Language Models
This addresses misalignments in information retrieval systems for users relying on LLM-generated expansions, but it is incremental as it builds on pseudo-relevance feedback methods.
The paper tackles the problem of query expansions from large language models (LLMs) misaligning with retrieval corpora, leading to hallucinations and outdated information, by introducing Corpus-Steered Query Expansion (CSQE) that incorporates corpus knowledge; experiments show it achieves strong performance without training, particularly for queries where LLMs lack knowledge.
Recent studies demonstrate that query expansions generated by large language models (LLMs) can considerably enhance information retrieval systems by generating hypothetical documents that answer the queries as expansions. However, challenges arise from misalignments between the expansions and the retrieval corpus, resulting in issues like hallucinations and outdated information due to the limited intrinsic knowledge of LLMs. Inspired by Pseudo Relevance Feedback (PRF), we introduce Corpus-Steered Query Expansion (CSQE) to promote the incorporation of knowledge embedded within the corpus. CSQE utilizes the relevance assessing capability of LLMs to systematically identify pivotal sentences in the initially-retrieved documents. These corpus-originated texts are subsequently used to expand the query together with LLM-knowledge empowered expansions, improving the relevance prediction between the query and the target documents. Extensive experiments reveal that CSQE exhibits strong performance without necessitating any training, especially with queries for which LLMs lack knowledge.