Rapidly Deploying a Neural Search Engine for the COVID-19 Open Research Dataset: Preliminary Thoughts and Lessons Learned
Edwin Zhang,1 Nikhil Gupta,1 Rodrigo Nogueira,1 Kyunghyun Cho,2,3,4,5 Jimmy Lin1 1 David R. Cheriton School of Computer Science, University of Waterloo 2 Courant Institute of Mathematical Sciences, New York University 3 Center for Data Science, New York University 4 Facebook AI Research 5 CIFAR Associate Fellow
Abstract
We present the Neural Covidex, a search engine that exploits the latest neural ranking architectures to provide information access to the COVID-19 Open Research Dataset curated by the Allen Institute for AI. This web application exists as part of a suite of tools that we have developed over the past few weeks to help domain experts tackle the ongoing global pandemic. We hope that improved information access capabilities to the scientific literature can inform evidence-based decision making and insight generation. This paper describes our initial efforts and offers a few thoughts about lessons we have learned along the way.
中文速览
面对新冠疫情暴发，如何让医生、公共卫生官员和科研人员快速从海量学术文献中找到可靠证据，是一个迫切的信息获取难题。研究团队在两周内基于 CORD-19 数据集（超过 4.7 万篇冠状病毒相关论文）构建了 Neural Covidex 搜索引擎，采用"多阶段排序（multi-stage ranking）"架构：先用 BM25 关键词检索召回候选段落，再用基于 T5 的序列到序列（sequence-to-sequence）神经重排序模型对结果精排，并借助 BioBERT 高亮与查询最相关的句子。系统不仅提供了供领域专家直接使用的网页搜索界面（含分面浏览功能），还将底层检索组件封装为可复用的开源工具包（Anserini/Pyserini），方便其他研究者在此基础上继续开发。该工作的意义在于，它在危机暴发的早期就将最先进的神经信息检索技术落地为实际可用的工具，为循证决策提供了更高质量的文献检索支持，同时也为如何在快速演变的紧急事件中敏捷构建 AI 工具提供了实践参考。
原文 arXiv:2004.05125;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2004.05125v1