@inproceedings{ma-etal-2023-learning,
title = "Learning {``}{O}{''} Helps for Learning More: Handling the Unlabeled Entity Problem for Class-incremental {NER}",
author = "Ma, Ruotian and
Chen, Xuanting and
Lin, Zhang and
Zhou, Xin and
Wang, Junzhe and
Gui, Tao and
Zhang, Qi and
Gao, Xiang and
Chen, Yun Wen",
editor = "Rogers, Anna and
Boyd-Graber, Jordan and
Okazaki, Naoaki",
booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.acl-long.328",
doi = "10.18653/v1/2023.acl-long.328",
pages = "5959--5979",
abstract = "As the categories of named entities rapidly increase, the deployed NER models are required to keep updating toward recognizing more entity types, creating a demand for class-incremental learning for NER. Considering the privacy concerns and storage constraints, the standard paradigm for class-incremental NER updates the models with training data only annotated with the new classes, yet the entities from other entity classes are regarded as {``}Non-entity{''} (or {``}O{''}). In this work, we conduct an empirical study on the {``}Unlabeled Entity Problem{''} and find that it leads to severe confusion between {``}O{''} and entities, decreasing class discrimination of old classes and declining the model{'}s ability to learn new classes. To solve the Unlabeled Entity Problem, we propose a novel representation learning method to learn discriminative representations for the entity classes and {``}O{''}. Specifically, we propose an entity-aware contrastive learning method that adaptively detects entity clusters in {``}O{''}. Furthermore, we propose two effective distance-based relabeling strategies for better learning the old classes. We introduce a more realistic and challenging benchmark for class-incremental NER, and the proposed method achieves up to 10.62{\%} improvement over the baseline methods.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ma-etal-2023-learning">
<titleInfo>
<title>Learning “O” Helps for Learning More: Handling the Unlabeled Entity Problem for Class-incremental NER</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ruotian</namePart>
<namePart type="family">Ma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuanting</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhang</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xin</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junzhe</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tao</namePart>
<namePart type="family">Gui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qi</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiang</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yun</namePart>
<namePart type="given">Wen</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anna</namePart>
<namePart type="family">Rogers</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jordan</namePart>
<namePart type="family">Boyd-Graber</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Naoaki</namePart>
<namePart type="family">Okazaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Toronto, Canada</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>As the categories of named entities rapidly increase, the deployed NER models are required to keep updating toward recognizing more entity types, creating a demand for class-incremental learning for NER. Considering the privacy concerns and storage constraints, the standard paradigm for class-incremental NER updates the models with training data only annotated with the new classes, yet the entities from other entity classes are regarded as “Non-entity” (or “O”). In this work, we conduct an empirical study on the “Unlabeled Entity Problem” and find that it leads to severe confusion between “O” and entities, decreasing class discrimination of old classes and declining the model’s ability to learn new classes. To solve the Unlabeled Entity Problem, we propose a novel representation learning method to learn discriminative representations for the entity classes and “O”. Specifically, we propose an entity-aware contrastive learning method that adaptively detects entity clusters in “O”. Furthermore, we propose two effective distance-based relabeling strategies for better learning the old classes. We introduce a more realistic and challenging benchmark for class-incremental NER, and the proposed method achieves up to 10.62% improvement over the baseline methods.</abstract>
<identifier type="citekey">ma-etal-2023-learning</identifier>
<identifier type="doi">10.18653/v1/2023.acl-long.328</identifier>
<location>
<url>https://aclanthology.org/2023.acl-long.328</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>5959</start>
<end>5979</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Learning “O” Helps for Learning More: Handling the Unlabeled Entity Problem for Class-incremental NER
%A Ma, Ruotian
%A Chen, Xuanting
%A Lin, Zhang
%A Zhou, Xin
%A Wang, Junzhe
%A Gui, Tao
%A Zhang, Qi
%A Gao, Xiang
%A Chen, Yun Wen
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F ma-etal-2023-learning
%X As the categories of named entities rapidly increase, the deployed NER models are required to keep updating toward recognizing more entity types, creating a demand for class-incremental learning for NER. Considering the privacy concerns and storage constraints, the standard paradigm for class-incremental NER updates the models with training data only annotated with the new classes, yet the entities from other entity classes are regarded as “Non-entity” (or “O”). In this work, we conduct an empirical study on the “Unlabeled Entity Problem” and find that it leads to severe confusion between “O” and entities, decreasing class discrimination of old classes and declining the model’s ability to learn new classes. To solve the Unlabeled Entity Problem, we propose a novel representation learning method to learn discriminative representations for the entity classes and “O”. Specifically, we propose an entity-aware contrastive learning method that adaptively detects entity clusters in “O”. Furthermore, we propose two effective distance-based relabeling strategies for better learning the old classes. We introduce a more realistic and challenging benchmark for class-incremental NER, and the proposed method achieves up to 10.62% improvement over the baseline methods.
%R 10.18653/v1/2023.acl-long.328
%U https://aclanthology.org/2023.acl-long.328
%U https://doi.org/10.18653/v1/2023.acl-long.328
%P 5959-5979
Markdown (Informal)
[Learning “O” Helps for Learning More: Handling the Unlabeled Entity Problem for Class-incremental NER](https://aclanthology.org/2023.acl-long.328) (Ma et al., ACL 2023)
ACL
- Ruotian Ma, Xuanting Chen, Zhang Lin, Xin Zhou, Junzhe Wang, Tao Gui, Qi Zhang, Xiang Gao, and Yun Wen Chen. 2023. Learning “O” Helps for Learning More: Handling the Unlabeled Entity Problem for Class-incremental NER. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5959–5979, Toronto, Canada. Association for Computational Linguistics.