NPRF: A Neural Pseudo Relevance Feedback Framework for Ad-hoc Information Retrieval
This work addresses the problem of query-document vocabulary mismatches in ad-hoc information retrieval for users of neural IR models, representing an incremental advancement by adapting PRF to neural architectures.
The paper tackles the challenge of integrating pseudo-relevance feedback (PRF) with neural information retrieval models by proposing an end-to-end neural PRF framework, which improves the performance of two state-of-the-art neural IR models as confirmed by experiments on standard test collections.
Pseudo-relevance feedback (PRF) is commonly used to boost the performance of traditional information retrieval (IR) models by using top-ranked documents to identify and weight new query terms, thereby reducing the effect of query-document vocabulary mismatches. While neural retrieval models have recently demonstrated strong results for ad-hoc retrieval, combining them with PRF is not straightforward due to incompatibilities between existing PRF approaches and neural architectures. To bridge this gap, we propose an end-to-end neural PRF framework that can be used with existing neural IR models by embedding different neural models as building blocks. Extensive experiments on two standard test collections confirm the effectiveness of the proposed NPRF framework in improving the performance of two state-of-the-art neural IR models.