A Little Bit Is Worse Than None: Ranking with Limited Training Data

Xinyu Zhang, Andrew Yates, Jimmy Lin


Abstract
Researchers have proposed simple yet effective techniques for the retrieval problem based on using BERT as a relevance classifier to rerank initial candidates from keyword search. In this work, we tackle the challenge of fine-tuning these models for specific domains in a data and computationally efficient manner. Typically, researchers fine-tune models using corpus-specific labeled data from sources such as TREC. We first answer the question: How much data of this type do we need? Recognizing that the most computationally efficient training is no training, we explore zero-shot ranking using BERT models that have already been fine-tuned with the large MS MARCO passage retrieval dataset. We arrive at the surprising and novel finding that “some” labeled in-domain data can be worse than none at all.
Anthology ID:
2020.sustainlp-1.14
Volume:
Proceedings of SustaiNLP: Workshop on Simple and Efficient Natural Language Processing
Month:
November
Year:
2020
Address:
Online
Venue:
sustainlp
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
107–112
Language:
URL:
https://aclanthology.org/2020.sustainlp-1.14
DOI:
10.18653/v1/2020.sustainlp-1.14
Bibkey:
Cite (ACL):
Xinyu Zhang, Andrew Yates, and Jimmy Lin. 2020. A Little Bit Is Worse Than None: Ranking with Limited Training Data. In Proceedings of SustaiNLP: Workshop on Simple and Efficient Natural Language Processing, pages 107–112, Online. Association for Computational Linguistics.
Cite (Informal):
A Little Bit Is Worse Than None: Ranking with Limited Training Data (Zhang et al., sustainlp 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/remove-xml-comments/2020.sustainlp-1.14.pdf
Video:
 https://slideslive.com/38939436
Data
MS MARCO