PGT: Pseudo Relevance Feedback Using a Graph-Based Transformer
This work addresses efficiency and accuracy challenges in information retrieval for researchers and practitioners, though it is incremental as it builds on existing Transformer and PRF methods.
The paper tackled the problem of integrating pseudo relevance feedback (PRF) into Transformer-based rerankers to improve accuracy while reducing computational costs, showing that PGT achieves at least as accurate results as full-attention Transformer PRF models with lower complexity.
Most research on pseudo relevance feedback (PRF) has been done in vector space and probabilistic retrieval models. This paper shows that Transformer-based rerankers can also benefit from the extra context that PRF provides. It presents PGT, a graph-based Transformer that sparsifies attention between graph nodes to enable PRF while avoiding the high computational complexity of most Transformer architectures. Experiments show that PGT improves upon non-PRF Transformer reranker, and it is at least as accurate as Transformer PRF models that use full attention, but with lower computational costs.