@inproceedings{zhou-etal-2022-textfusion,
title = "{T}ext{F}usion: Privacy-Preserving Pre-trained Model Inference via Token Fusion",
author = "Zhou, Xin and
Lu, Jinzhu and
Gui, Tao and
Ma, Ruotian and
Fei, Zichu and
Wang, Yuran and
Ding, Yong and
Cheung, Yibo and
Zhang, Qi and
Huang, Xuanjing",
editor = "Goldberg, Yoav and
Kozareva, Zornitsa and
Zhang, Yue",
booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2022",
address = "Abu Dhabi, United Arab Emirates",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.emnlp-main.572/",
doi = "10.18653/v1/2022.emnlp-main.572",
pages = "8360--8371",
abstract = "Recently, more and more pre-trained language models are released as a cloud service. It allows users who lack computing resources to perform inference with a powerful model by uploading data to the cloud. The plain text may contain private information, as the result, users prefer to do partial computations locally and upload intermediate representations to the cloud for subsequent inference.However, recent studies have shown that intermediate representations can also be recovered to plain text with reasonable accuracy, thus the risk of privacy leakage still exists. To address this issue, we propose TextFusion, a novel method for preserving inference privacy.Specifically, we train a Fusion Predictor to dynamically fuse token representations, which hides multiple private token representations behind an unrecognizable one.Furthermore, an adversarial training regime is employed to privatize these representations. In this way, the cloud only receives incomplete and perturbed representations, making it difficult to accurately recover the complete plain text.The experimental results on diverse classification tasks show that our approach can effectively preserve inference privacy without significantly sacrificing performance in different scenarios."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhou-etal-2022-textfusion">
<titleInfo>
<title>TextFusion: Privacy-Preserving Pre-trained Model Inference via Token Fusion</title>
</titleInfo>
<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">Jinzhu</namePart>
<namePart type="family">Lu</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">Ruotian</namePart>
<namePart type="family">Ma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zichu</namePart>
<namePart type="family">Fei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuran</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yong</namePart>
<namePart type="family">Ding</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yibo</namePart>
<namePart type="family">Cheung</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">Xuanjing</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yoav</namePart>
<namePart type="family">Goldberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zornitsa</namePart>
<namePart type="family">Kozareva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yue</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">Abu Dhabi, United Arab Emirates</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Recently, more and more pre-trained language models are released as a cloud service. It allows users who lack computing resources to perform inference with a powerful model by uploading data to the cloud. The plain text may contain private information, as the result, users prefer to do partial computations locally and upload intermediate representations to the cloud for subsequent inference.However, recent studies have shown that intermediate representations can also be recovered to plain text with reasonable accuracy, thus the risk of privacy leakage still exists. To address this issue, we propose TextFusion, a novel method for preserving inference privacy.Specifically, we train a Fusion Predictor to dynamically fuse token representations, which hides multiple private token representations behind an unrecognizable one.Furthermore, an adversarial training regime is employed to privatize these representations. In this way, the cloud only receives incomplete and perturbed representations, making it difficult to accurately recover the complete plain text.The experimental results on diverse classification tasks show that our approach can effectively preserve inference privacy without significantly sacrificing performance in different scenarios.</abstract>
<identifier type="citekey">zhou-etal-2022-textfusion</identifier>
<identifier type="doi">10.18653/v1/2022.emnlp-main.572</identifier>
<location>
<url>https://aclanthology.org/2022.emnlp-main.572/</url>
</location>
<part>
<date>2022-12</date>
<extent unit="page">
<start>8360</start>
<end>8371</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T TextFusion: Privacy-Preserving Pre-trained Model Inference via Token Fusion
%A Zhou, Xin
%A Lu, Jinzhu
%A Gui, Tao
%A Ma, Ruotian
%A Fei, Zichu
%A Wang, Yuran
%A Ding, Yong
%A Cheung, Yibo
%A Zhang, Qi
%A Huang, Xuanjing
%Y Goldberg, Yoav
%Y Kozareva, Zornitsa
%Y Zhang, Yue
%S Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
%D 2022
%8 December
%I Association for Computational Linguistics
%C Abu Dhabi, United Arab Emirates
%F zhou-etal-2022-textfusion
%X Recently, more and more pre-trained language models are released as a cloud service. It allows users who lack computing resources to perform inference with a powerful model by uploading data to the cloud. The plain text may contain private information, as the result, users prefer to do partial computations locally and upload intermediate representations to the cloud for subsequent inference.However, recent studies have shown that intermediate representations can also be recovered to plain text with reasonable accuracy, thus the risk of privacy leakage still exists. To address this issue, we propose TextFusion, a novel method for preserving inference privacy.Specifically, we train a Fusion Predictor to dynamically fuse token representations, which hides multiple private token representations behind an unrecognizable one.Furthermore, an adversarial training regime is employed to privatize these representations. In this way, the cloud only receives incomplete and perturbed representations, making it difficult to accurately recover the complete plain text.The experimental results on diverse classification tasks show that our approach can effectively preserve inference privacy without significantly sacrificing performance in different scenarios.
%R 10.18653/v1/2022.emnlp-main.572
%U https://aclanthology.org/2022.emnlp-main.572/
%U https://doi.org/10.18653/v1/2022.emnlp-main.572
%P 8360-8371
Markdown (Informal)
[TextFusion: Privacy-Preserving Pre-trained Model Inference via Token Fusion](https://aclanthology.org/2022.emnlp-main.572/) (Zhou et al., EMNLP 2022)
ACL
- Xin Zhou, Jinzhu Lu, Tao Gui, Ruotian Ma, Zichu Fei, Yuran Wang, Yong Ding, Yibo Cheung, Qi Zhang, and Xuanjing Huang. 2022. TextFusion: Privacy-Preserving Pre-trained Model Inference via Token Fusion. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 8360–8371, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.