SLEDGE-Z: A Zero-Shot Baseline for COVID-19 Literature Search

TL;DR

SLEDGE-Z introduces a zero-shot COVID-19 literature search method that leverages scientific pre-training and data filtering, achieving top performance without relying on COVID-specific training data.

Contribution

The paper presents a novel zero-shot ranking algorithm for COVID-19 literature search that outperforms existing models and sets a new baseline for rapid, effective scientific article retrieval.

Findings

Achieves P@5 of 0.80 and nDCG@10 of 0.68 on TREC COVID benchmarks

Outperforms models trained specifically on COVID data despite no such training

Ranks among the top zero-shot methods on the TREC COVID leaderboard

Abstract

With worldwide concerns surrounding the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), there is a rapidly growing body of scientific literature on the virus. Clinicians, researchers, and policy-makers need to be able to search these articles effectively. In this work, we present a zero-shot ranking algorithm that adapts to COVID-related scientific literature. Our approach filters training data from another collection down to medical-related queries, uses a neural re-ranking model pre-trained on scientific text (SciBERT), and filters the target document collection. This approach ranks top among zero-shot methods on the TREC COVID Round 1 leaderboard, and exhibits a P@5 of 0.80 and an nDCG@10 of 0.68 when evaluated on both Round 1 and 2 judgments. Despite not relying on TREC-COVID data, our method outperforms models that do. As one of the first search methods to thoroughly evaluate COVID-19 search, we hope that this serves as a strong baseline and helps in the global crisis.