FABERT: A Feature Aggregation BERT-Based Model for Document Reranking

Xiaozhi Zhu, Leung Pun Wong, Lap Kei Lee, Hai Liu, Tianyong Hao

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In a document reranking task, pre-trained language models such as BERT have been successfully applied due to their powerful capability in extracting informative features from queries and candidate answers. However, these language models always generate discriminative features and pay less attention to generalized features which contain shared information of query-answer pairs to assist question answering. In this paper, we propose a BERT-based model named FABERT by integrating both discriminative features and generalized features produced by a gradient reverse layer into one answer vector with an attention mechanism for document reranking. Extensive experiments on the MS MARCO passage ranking task and TREC Robust dataset show that FABERT outperforms baseline methods including a feature projection method which projects existing feature vectors into the orthogonal space of generalized feature vector to eliminate common information of generalized feature vectors.

Original languageEnglish
Title of host publicationNatural Language Processing and Chinese Computing - 10th CCF International Conference, NLPCC 2021, Proceedings
EditorsLu Wang, Yansong Feng, Yu Hong, Ruifang He
Pages139-150
Number of pages12
DOIs
Publication statusPublished - 2021
Event10th CCF Conference on Natural Language Processing and Chinese Computing, NLPCC 2021 - Qingdao, China
Duration: 13 Oct 202117 Oct 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13029 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference10th CCF Conference on Natural Language Processing and Chinese Computing, NLPCC 2021
Country/TerritoryChina
CityQingdao
Period13/10/2117/10/21

Keywords

  • BERT
  • Document ranking
  • Feature aggregation

Fingerprint

Dive into the research topics of 'FABERT: A Feature Aggregation BERT-Based Model for Document Reranking'. Together they form a unique fingerprint.

Cite this