@inproceedings{basu-etal-2022-strategies,
title = "Strategies to Improve Few-shot Learning for Intent Classification and Slot-Filling",
author = "Basu, Samyadeep and
Sharaf, Amr and
Ip Kiun Chong, Karine and
Fischer, Alex and
Rohra, Vishal and
Amoake, Michael and
El-Hammamy, Hazem and
Nosakhare, Ehi and
Ramani, Vijay and
Han, Benjamin",
editor = "Chen, Wenhu and
Chen, Xinyun and
Chen, Zhiyu and
Yao, Ziyu and
Yasunaga, Michihiro and
Yu, Tao and
Zhang, Rui",
booktitle = "Proceedings of the Workshop on Structured and Unstructured Knowledge Integration (SUKI)",
month = jul,
year = "2022",
address = "Seattle, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.suki-1.3",
doi = "10.18653/v1/2022.suki-1.3",
pages = "17--25",
abstract = "Intent classification (IC) and slot filling (SF) are two fundamental tasks in modern Natural Language Understanding (NLU) systems. Collecting and annotating large amounts of data to train deep learning models for such systems are not scalable. This problem can be addressed by learning from few examples using fast supervised meta-learning techniques such as prototypical networks. In this work, we systematically investigate how contrastive learning and data augmentation methods can benefit these existing meta-learning pipelines for jointly modelled IC/SF tasks. Through extensive experiments across standard IC/SF benchmarks (SNIPS and ATIS), we show that our proposed approaches outperform standard meta-learning methods: contrastive losses as a regularizer in conjunction with prototypical networks consistently outperform the existing state-of-the-art for both IC and SF tasks, while data augmentation strategies primarily improve few-shot IC by a significant margin",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="basu-etal-2022-strategies">
<titleInfo>
<title>Strategies to Improve Few-shot Learning for Intent Classification and Slot-Filling</title>
</titleInfo>
<name type="personal">
<namePart type="given">Samyadeep</namePart>
<namePart type="family">Basu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Amr</namePart>
<namePart type="family">Sharaf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Karine</namePart>
<namePart type="family">Ip Kiun Chong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alex</namePart>
<namePart type="family">Fischer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vishal</namePart>
<namePart type="family">Rohra</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Michael</namePart>
<namePart type="family">Amoake</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hazem</namePart>
<namePart type="family">El-Hammamy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ehi</namePart>
<namePart type="family">Nosakhare</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vijay</namePart>
<namePart type="family">Ramani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Benjamin</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Workshop on Structured and Unstructured Knowledge Integration (SUKI)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wenhu</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinyun</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhiyu</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ziyu</namePart>
<namePart type="family">Yao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Michihiro</namePart>
<namePart type="family">Yasunaga</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tao</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rui</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Seattle, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Intent classification (IC) and slot filling (SF) are two fundamental tasks in modern Natural Language Understanding (NLU) systems. Collecting and annotating large amounts of data to train deep learning models for such systems are not scalable. This problem can be addressed by learning from few examples using fast supervised meta-learning techniques such as prototypical networks. In this work, we systematically investigate how contrastive learning and data augmentation methods can benefit these existing meta-learning pipelines for jointly modelled IC/SF tasks. Through extensive experiments across standard IC/SF benchmarks (SNIPS and ATIS), we show that our proposed approaches outperform standard meta-learning methods: contrastive losses as a regularizer in conjunction with prototypical networks consistently outperform the existing state-of-the-art for both IC and SF tasks, while data augmentation strategies primarily improve few-shot IC by a significant margin</abstract>
<identifier type="citekey">basu-etal-2022-strategies</identifier>
<identifier type="doi">10.18653/v1/2022.suki-1.3</identifier>
<location>
<url>https://aclanthology.org/2022.suki-1.3</url>
</location>
<part>
<date>2022-07</date>
<extent unit="page">
<start>17</start>
<end>25</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Strategies to Improve Few-shot Learning for Intent Classification and Slot-Filling
%A Basu, Samyadeep
%A Sharaf, Amr
%A Ip Kiun Chong, Karine
%A Fischer, Alex
%A Rohra, Vishal
%A Amoake, Michael
%A El-Hammamy, Hazem
%A Nosakhare, Ehi
%A Ramani, Vijay
%A Han, Benjamin
%Y Chen, Wenhu
%Y Chen, Xinyun
%Y Chen, Zhiyu
%Y Yao, Ziyu
%Y Yasunaga, Michihiro
%Y Yu, Tao
%Y Zhang, Rui
%S Proceedings of the Workshop on Structured and Unstructured Knowledge Integration (SUKI)
%D 2022
%8 July
%I Association for Computational Linguistics
%C Seattle, USA
%F basu-etal-2022-strategies
%X Intent classification (IC) and slot filling (SF) are two fundamental tasks in modern Natural Language Understanding (NLU) systems. Collecting and annotating large amounts of data to train deep learning models for such systems are not scalable. This problem can be addressed by learning from few examples using fast supervised meta-learning techniques such as prototypical networks. In this work, we systematically investigate how contrastive learning and data augmentation methods can benefit these existing meta-learning pipelines for jointly modelled IC/SF tasks. Through extensive experiments across standard IC/SF benchmarks (SNIPS and ATIS), we show that our proposed approaches outperform standard meta-learning methods: contrastive losses as a regularizer in conjunction with prototypical networks consistently outperform the existing state-of-the-art for both IC and SF tasks, while data augmentation strategies primarily improve few-shot IC by a significant margin
%R 10.18653/v1/2022.suki-1.3
%U https://aclanthology.org/2022.suki-1.3
%U https://doi.org/10.18653/v1/2022.suki-1.3
%P 17-25
Markdown (Informal)
[Strategies to Improve Few-shot Learning for Intent Classification and Slot-Filling](https://aclanthology.org/2022.suki-1.3) (Basu et al., SUKI 2022)
ACL
- Samyadeep Basu, Amr Sharaf, Karine Ip Kiun Chong, Alex Fischer, Vishal Rohra, Michael Amoake, Hazem El-Hammamy, Ehi Nosakhare, Vijay Ramani, and Benjamin Han. 2022. Strategies to Improve Few-shot Learning for Intent Classification and Slot-Filling. In Proceedings of the Workshop on Structured and Unstructured Knowledge Integration (SUKI), pages 17–25, Seattle, USA. Association for Computational Linguistics.