@inproceedings{bhattacharya-etal-2023-cupid,
title = "{CUPID}: Curriculum Learning Based Real-Time Prediction using Distillation",
author = "Bhattacharya, Arindam and
Ms, Ankith and
Gandhi, Ankit and
Huddar, Vijay and
Saroop, Atul and
Bhagat, Rahul",
editor = "Sitaram, Sunayana and
Beigman Klebanov, Beata and
Williams, Jason D",
booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 5: Industry Track)",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.acl-industry.69",
doi = "10.18653/v1/2023.acl-industry.69",
pages = "720--728",
abstract = "Relevance in E-commerce Product Search is crucial for providing customers with accurate results that match their query intent. With recent advancements in NLP and Deep Learning, Transformers have become the default choice for relevance classification tasks. In such a setting, the relevance model uses query text and product title as input features, and estimates if the product is relevant for the customer query. While cross-attention in Transformers enables a more accurate relevance prediction in such a setting, its high evaluation latency makes it unsuitable for real-time predictions in which thousands of products must be evaluated against a user query within few milliseconds. To address this issue, we propose CUPID: a Curriculum learning based real-time Prediction using Distillation that utilizes knowledge distillation within a curriculum learning setting to learn a simpler architecture that can be evaluated within low latency budgets. In a bi-lingual relevance prediction task, our approach shows an 302 bps improvement on English and 676 bps improvement for low-resource Arabic, while maintaining the low evaluation latency on CPUs.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bhattacharya-etal-2023-cupid">
<titleInfo>
<title>CUPID: Curriculum Learning Based Real-Time Prediction using Distillation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Arindam</namePart>
<namePart type="family">Bhattacharya</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ankith</namePart>
<namePart type="family">Ms</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ankit</namePart>
<namePart type="family">Gandhi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vijay</namePart>
<namePart type="family">Huddar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Atul</namePart>
<namePart type="family">Saroop</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rahul</namePart>
<namePart type="family">Bhagat</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 5: Industry Track)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sunayana</namePart>
<namePart type="family">Sitaram</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Beata</namePart>
<namePart type="family">Beigman Klebanov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jason</namePart>
<namePart type="given">D</namePart>
<namePart type="family">Williams</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>Relevance in E-commerce Product Search is crucial for providing customers with accurate results that match their query intent. With recent advancements in NLP and Deep Learning, Transformers have become the default choice for relevance classification tasks. In such a setting, the relevance model uses query text and product title as input features, and estimates if the product is relevant for the customer query. While cross-attention in Transformers enables a more accurate relevance prediction in such a setting, its high evaluation latency makes it unsuitable for real-time predictions in which thousands of products must be evaluated against a user query within few milliseconds. To address this issue, we propose CUPID: a Curriculum learning based real-time Prediction using Distillation that utilizes knowledge distillation within a curriculum learning setting to learn a simpler architecture that can be evaluated within low latency budgets. In a bi-lingual relevance prediction task, our approach shows an 302 bps improvement on English and 676 bps improvement for low-resource Arabic, while maintaining the low evaluation latency on CPUs.</abstract>
<identifier type="citekey">bhattacharya-etal-2023-cupid</identifier>
<identifier type="doi">10.18653/v1/2023.acl-industry.69</identifier>
<location>
<url>https://aclanthology.org/2023.acl-industry.69</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>720</start>
<end>728</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CUPID: Curriculum Learning Based Real-Time Prediction using Distillation
%A Bhattacharya, Arindam
%A Ms, Ankith
%A Gandhi, Ankit
%A Huddar, Vijay
%A Saroop, Atul
%A Bhagat, Rahul
%Y Sitaram, Sunayana
%Y Beigman Klebanov, Beata
%Y Williams, Jason D.
%S Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 5: Industry Track)
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F bhattacharya-etal-2023-cupid
%X Relevance in E-commerce Product Search is crucial for providing customers with accurate results that match their query intent. With recent advancements in NLP and Deep Learning, Transformers have become the default choice for relevance classification tasks. In such a setting, the relevance model uses query text and product title as input features, and estimates if the product is relevant for the customer query. While cross-attention in Transformers enables a more accurate relevance prediction in such a setting, its high evaluation latency makes it unsuitable for real-time predictions in which thousands of products must be evaluated against a user query within few milliseconds. To address this issue, we propose CUPID: a Curriculum learning based real-time Prediction using Distillation that utilizes knowledge distillation within a curriculum learning setting to learn a simpler architecture that can be evaluated within low latency budgets. In a bi-lingual relevance prediction task, our approach shows an 302 bps improvement on English and 676 bps improvement for low-resource Arabic, while maintaining the low evaluation latency on CPUs.
%R 10.18653/v1/2023.acl-industry.69
%U https://aclanthology.org/2023.acl-industry.69
%U https://doi.org/10.18653/v1/2023.acl-industry.69
%P 720-728
Markdown (Informal)
[CUPID: Curriculum Learning Based Real-Time Prediction using Distillation](https://aclanthology.org/2023.acl-industry.69) (Bhattacharya et al., ACL 2023)
ACL
- Arindam Bhattacharya, Ankith Ms, Ankit Gandhi, Vijay Huddar, Atul Saroop, and Rahul Bhagat. 2023. CUPID: Curriculum Learning Based Real-Time Prediction using Distillation. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 5: Industry Track), pages 720–728, Toronto, Canada. Association for Computational Linguistics.