Lightweight Composite Re-Ranking for Efficient Keyword Search with BERT
Lightweight Composite Re-Ranking for Efficient Keyword Search with BERT
复制标题
DOI：
10.1145/3488560.3498495
复制
发表时间：
2021-03
期刊：
影响因子：
--
通讯作者：
Yingrui Yang;Yifan Qiao;Jinjin Shao;Xifeng Yan;Tao Yang
中科院分区：
文献类型：
--
作者：
Yingrui Yang;Yifan Qiao;Jinjin Shao;Xifeng Yan;Tao Yang
最近，基于transformer的排名模型已被证明可以为文档搜索提供高相关性，并且相关性-效率权衡对于快速查询响应时间变得重要。本文提出了BECR（BERT-based Composite Re-Ranking），这是一种轻量级的复合重排序方案，它结合了深度上下文令牌交互和传统的词汇术语匹配功能。BECR进行查询分解，并使用基于uni-gram和skip-n-gram的可预计算的令牌嵌入来组成查询表示，以寻求推理效率和相关性的折衷。因此，它不执行昂贵的Transformer计算在线推理过程中，不需要使用GPU。本文描述了一个相关性和效率的BECR与几个TREC数据集的评估。
Recently transformer-based ranking models have been shown to deliver high relevance for document search and the relevance-efficiency tradeoff becomes important for fast query response times. This paper presents BECR (BERT-based Composite Re-Ranking), a lightweight composite re-ranking scheme that combines deep contextual token interactions and traditional lexical term-matching features. BECR conducts query decomposition and composes a query representation using pre-computable token embeddings based on uni-grams and skip-n-grams, to seek a tradeoff of inference efficiency and relevance. Thus it does not perform expensive transformer computations during online inference, and does not require the use of GPU. This paper describes an evaluation of relevance and efficiency of BECR with several TREC datasets.