@inproceedings{zeng-etal-2020-tri,
title = "Tri-Train: Automatic Pre-Fine Tuning between Pre-Training and Fine-Tuning for {S}ci{NER}",
author = "Zeng, Qingkai and
Yu, Wenhao and
Yu, Mengxia and
Jiang, Tianwen and
Weninger, Tim and
Jiang, Meng",
editor = "Cohn, Trevor and
He, Yulan and
Liu, Yang",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
month = nov,
year = "2020",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2020.findings-emnlp.429",
doi = "10.18653/v1/2020.findings-emnlp.429",
pages = "4778--4787",
abstract = "The training process of scientific NER models is commonly performed in two steps: i) Pre-training a language model by self-supervised tasks on huge data and ii) fine-tune training with small labelled data. The success of the strategy depends on the relevance between the data domains and between the tasks. However, gaps are found in practice when the target domains are specific and small. We propose a novel framework to introduce a {``}pre-fine tuning{''} step between pre-training and fine-tuning. It constructs a corpus by selecting sentences from unlabeled documents that are the most relevant with the labelled training data. Instead of predicting tokens in random spans, the pre-fine tuning task is to predict tokens in entity candidates identified by text mining methods. Pre-fine tuning is automatic and light-weight because the corpus size can be much smaller than pre-training data to achieve a better performance. Experiments on seven benchmarks demonstrate the effectiveness.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zeng-etal-2020-tri">
<titleInfo>
<title>Tri-Train: Automatic Pre-Fine Tuning between Pre-Training and Fine-Tuning for SciNER</title>
</titleInfo>
<name type="personal">
<namePart type="given">Qingkai</namePart>
<namePart type="family">Zeng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wenhao</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mengxia</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianwen</namePart>
<namePart type="family">Jiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tim</namePart>
<namePart type="family">Weninger</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Meng</namePart>
<namePart type="family">Jiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2020-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2020</title>
</titleInfo>
<name type="personal">
<namePart type="given">Trevor</namePart>
<namePart type="family">Cohn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yulan</namePart>
<namePart type="family">He</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>The training process of scientific NER models is commonly performed in two steps: i) Pre-training a language model by self-supervised tasks on huge data and ii) fine-tune training with small labelled data. The success of the strategy depends on the relevance between the data domains and between the tasks. However, gaps are found in practice when the target domains are specific and small. We propose a novel framework to introduce a “pre-fine tuning” step between pre-training and fine-tuning. It constructs a corpus by selecting sentences from unlabeled documents that are the most relevant with the labelled training data. Instead of predicting tokens in random spans, the pre-fine tuning task is to predict tokens in entity candidates identified by text mining methods. Pre-fine tuning is automatic and light-weight because the corpus size can be much smaller than pre-training data to achieve a better performance. Experiments on seven benchmarks demonstrate the effectiveness.</abstract>
<identifier type="citekey">zeng-etal-2020-tri</identifier>
<identifier type="doi">10.18653/v1/2020.findings-emnlp.429</identifier>
<location>
<url>https://aclanthology.org/2020.findings-emnlp.429</url>
</location>
<part>
<date>2020-11</date>
<extent unit="page">
<start>4778</start>
<end>4787</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Tri-Train: Automatic Pre-Fine Tuning between Pre-Training and Fine-Tuning for SciNER
%A Zeng, Qingkai
%A Yu, Wenhao
%A Yu, Mengxia
%A Jiang, Tianwen
%A Weninger, Tim
%A Jiang, Meng
%Y Cohn, Trevor
%Y He, Yulan
%Y Liu, Yang
%S Findings of the Association for Computational Linguistics: EMNLP 2020
%D 2020
%8 November
%I Association for Computational Linguistics
%C Online
%F zeng-etal-2020-tri
%X The training process of scientific NER models is commonly performed in two steps: i) Pre-training a language model by self-supervised tasks on huge data and ii) fine-tune training with small labelled data. The success of the strategy depends on the relevance between the data domains and between the tasks. However, gaps are found in practice when the target domains are specific and small. We propose a novel framework to introduce a “pre-fine tuning” step between pre-training and fine-tuning. It constructs a corpus by selecting sentences from unlabeled documents that are the most relevant with the labelled training data. Instead of predicting tokens in random spans, the pre-fine tuning task is to predict tokens in entity candidates identified by text mining methods. Pre-fine tuning is automatic and light-weight because the corpus size can be much smaller than pre-training data to achieve a better performance. Experiments on seven benchmarks demonstrate the effectiveness.
%R 10.18653/v1/2020.findings-emnlp.429
%U https://aclanthology.org/2020.findings-emnlp.429
%U https://doi.org/10.18653/v1/2020.findings-emnlp.429
%P 4778-4787
Markdown (Informal)
[Tri-Train: Automatic Pre-Fine Tuning between Pre-Training and Fine-Tuning for SciNER](https://aclanthology.org/2020.findings-emnlp.429) (Zeng et al., Findings 2020)
ACL