MatRank: Text Re-ranking by Latent Preference Matrix

Jinwen Luo, Jiuding Yang, Weidong Guo, Chenglin Li, Di Niu, Yu Xu


Abstract
Text ranking plays a key role in providing content that best answers user queries. It is usually divided into two sub-tasks to perform efficient information retrieval given a query: text retrieval and text re-ranking. Recent research on pretrained language models (PLM) has demonstrated efficiency and gain on both sub-tasks. However, while existing methods have benefited from pre-trained language models and achieved high recall rates on passage retrieval, the ranking performance still demands further improvement. In this paper, we propose MatRank, which learns to re-rank the text retrieved for a given query by learning to predict the most relevant passage based on a latent preference matrix. Specifically, MatRank uses a PLM to generate an asymmetric latent matrix of relative preference scores between all pairs of retrieved passages. Then, the latent matrix is aggregated row-wise and column-wise to obtain global preferences and predictions of the most relevant passage in two of these directions, respectively. We conduct extensive experiments on MS MACRO, WikiAQ, and SemEval datasets. Experimental results show that MatRank has achieved new state-of-the-art results on these datasets, outperforming all prior methods on ranking performance metrics.
Anthology ID:
2022.findings-emnlp.146
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2022
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2011–2023
Language:
URL:
https://aclanthology.org/2022.findings-emnlp.146
DOI:
10.18653/v1/2022.findings-emnlp.146
Bibkey:
Cite (ACL):
Jinwen Luo, Jiuding Yang, Weidong Guo, Chenglin Li, Di Niu, and Yu Xu. 2022. MatRank: Text Re-ranking by Latent Preference Matrix. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 2011–2023, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
MatRank: Text Re-ranking by Latent Preference Matrix (Luo et al., Findings 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.findings-emnlp.146.pdf
Software:
 2022.findings-emnlp.146.software.zip
Video:
 https://aclanthology.org/2022.findings-emnlp.146.mp4