End-to-End Retrieval in Continuous Space
Daniel Gillick, Alessandro Presta, Gaurav Singh Tomar Google AI {dgillick, apresta,
Abstract
Most text-based information retrieval (IR) systems index objects by words or phrases. These discrete systems have been augmented by models that use embeddings to measure similarity in continuous space. But continuous-space models are typically used just to re-rank the top candidates. We consider the problem of end-to-end continuous retrieval, where standard approximate nearest neighbor (ANN) search replaces the usual discrete inverted index, and rely entirely on distances between learned embeddings. By training simple models specifically for retrieval, with an appropriate model architecture, we improve on a discrete baseline by 8% and 26% (MAP) on two similar-question retrieval tasks. We also discuss the problem of evaluation for retrieval systems, and show how to modify existing pairwise similarity datasets for this purpose.
中文速览
传统信息检索系统依赖离散的倒排索引，只能靠词语匹配来找文档，而基于语义向量（embedding）的模型通常只被用来对候选结果重新排序，并非直接检索。本文提出了一种"端到端连续检索"方案：用双编码器（Dual Encoder）将查询和候选文本分别编码为稠密向量，再直接通过近似最近邻（ANN）搜索完成检索，完全绕开倒排索引。研究者还提出了一种叫作"批内采样softmax（in-batch sampled softmax）"的负样本训练策略，并将现有成对相似度数据集改造成完整的检索评测集。在Quora相似问题和AskUbuntu问题检索两项任务上，该方法的MAP指标分别比BM25离散检索基线提升了8%和26%，证明专门为检索任务训练的简单语义向量模型可以显著超越依赖词匹配的传统方法，为大规模语义检索提供了有力的实证支撑。
原文 arXiv:1811.08008;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1811.08008v1