PASH at TREC 2021 Deep Learning Track: Generative Enhanced Model for Multi-stage Ranking

TL;DR

This paper presents the PASH system's participation in TREC 2021, utilizing a multi-stage ranking approach with generative models to improve retrieval performance in deep learning-based information retrieval tasks.

Contribution

The paper introduces a novel multi-stage ranking framework combining sparse, dense, and generative models, notably integrating T5 to enhance retrieval effectiveness.

Findings

Improved retrieval performance over previous TREC 2020 methods.

Effective combination of sparse, dense, and generative models.

Enhanced ranking accuracy with T5 integration.

Abstract

This paper describes the PASH participation in TREC 2021 Deep Learning Track. In the recall stage, we adopt a scheme combining sparse and dense retrieval method. In the multi-stage ranking phase, point-wise and pair-wise ranking strategies are used one after another based on model continual pre-trained on general knowledge and document-level data. Compared to TREC 2020 Deep Learning Track, we have additionally introduced the generative model T5 to further enhance the performance.