@inproceedings{elmahdy-salem-2024-deconstructing,
title = "Deconstructing Classifiers: Towards A Data Reconstruction Attack Against Text Classification Models",
author = "Elmahdy, Adel and
Salem, Ahmed",
editor = "Habernal, Ivan and
Ghanavati, Sepideh and
Ravichander, Abhilasha and
Jain, Vijayanta and
Thaine, Patricia and
Igamberdiev, Timour and
Mireshghallah, Niloofar and
Feyisetan, Oluwaseyi",
booktitle = "Proceedings of the Fifth Workshop on Privacy in Natural Language Processing",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.privatenlp-1.15",
pages = "143--158",
abstract = "Natural language processing (NLP) models have become increasingly popular in real-world applications, such as text classification. However, they are vulnerable to privacy attacks, including data reconstruction attacks that aim to extract the data used to train the model. Most previous studies on data reconstruction attacks have focused on LLM, while classification models were assumed to be more secure. In this work, we propose a new targeted data reconstruction attack called the Mix And Match attack, which takes advantage of the fact that most classification models are based on LLM. The Mix And Match attack uses the base model of the target model to generate candidate tokens and then prunes them using the classification head. We extensively demonstrate the effectiveness of the attack using both random and organic canaries. This work highlights the importance of considering the privacy risks associated with data reconstruction attacks in classification models and offers insights into possible leakages.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="elmahdy-salem-2024-deconstructing">
<titleInfo>
<title>Deconstructing Classifiers: Towards A Data Reconstruction Attack Against Text Classification Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Adel</namePart>
<namePart type="family">Elmahdy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ahmed</namePart>
<namePart type="family">Salem</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fifth Workshop on Privacy in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ivan</namePart>
<namePart type="family">Habernal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sepideh</namePart>
<namePart type="family">Ghanavati</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abhilasha</namePart>
<namePart type="family">Ravichander</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vijayanta</namePart>
<namePart type="family">Jain</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Patricia</namePart>
<namePart type="family">Thaine</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Timour</namePart>
<namePart type="family">Igamberdiev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Niloofar</namePart>
<namePart type="family">Mireshghallah</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Oluwaseyi</namePart>
<namePart type="family">Feyisetan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Natural language processing (NLP) models have become increasingly popular in real-world applications, such as text classification. However, they are vulnerable to privacy attacks, including data reconstruction attacks that aim to extract the data used to train the model. Most previous studies on data reconstruction attacks have focused on LLM, while classification models were assumed to be more secure. In this work, we propose a new targeted data reconstruction attack called the Mix And Match attack, which takes advantage of the fact that most classification models are based on LLM. The Mix And Match attack uses the base model of the target model to generate candidate tokens and then prunes them using the classification head. We extensively demonstrate the effectiveness of the attack using both random and organic canaries. This work highlights the importance of considering the privacy risks associated with data reconstruction attacks in classification models and offers insights into possible leakages.</abstract>
<identifier type="citekey">elmahdy-salem-2024-deconstructing</identifier>
<location>
<url>https://aclanthology.org/2024.privatenlp-1.15</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>143</start>
<end>158</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Deconstructing Classifiers: Towards A Data Reconstruction Attack Against Text Classification Models
%A Elmahdy, Adel
%A Salem, Ahmed
%Y Habernal, Ivan
%Y Ghanavati, Sepideh
%Y Ravichander, Abhilasha
%Y Jain, Vijayanta
%Y Thaine, Patricia
%Y Igamberdiev, Timour
%Y Mireshghallah, Niloofar
%Y Feyisetan, Oluwaseyi
%S Proceedings of the Fifth Workshop on Privacy in Natural Language Processing
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand
%F elmahdy-salem-2024-deconstructing
%X Natural language processing (NLP) models have become increasingly popular in real-world applications, such as text classification. However, they are vulnerable to privacy attacks, including data reconstruction attacks that aim to extract the data used to train the model. Most previous studies on data reconstruction attacks have focused on LLM, while classification models were assumed to be more secure. In this work, we propose a new targeted data reconstruction attack called the Mix And Match attack, which takes advantage of the fact that most classification models are based on LLM. The Mix And Match attack uses the base model of the target model to generate candidate tokens and then prunes them using the classification head. We extensively demonstrate the effectiveness of the attack using both random and organic canaries. This work highlights the importance of considering the privacy risks associated with data reconstruction attacks in classification models and offers insights into possible leakages.
%U https://aclanthology.org/2024.privatenlp-1.15
%P 143-158
Markdown (Informal)
[Deconstructing Classifiers: Towards A Data Reconstruction Attack Against Text Classification Models](https://aclanthology.org/2024.privatenlp-1.15) (Elmahdy & Salem, PrivateNLP-WS 2024)
ACL