@inproceedings{bhan-etal-2023-enhancing,
title = "Enhancing textual counterfactual explanation intelligibility through Counterfactual Feature Importance",
author = "Bhan, Milan and
Vittaut, Jean-noel and
Chesneau, Nicolas and
Lesot, Marie-jeanne",
editor = "Ovalle, Anaelia and
Chang, Kai-Wei and
Mehrabi, Ninareh and
Pruksachatkun, Yada and
Galystan, Aram and
Dhamala, Jwala and
Verma, Apurv and
Cao, Trista and
Kumar, Anoop and
Gupta, Rahul",
booktitle = "Proceedings of the 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023)",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.trustnlp-1.19",
doi = "10.18653/v1/2023.trustnlp-1.19",
pages = "221--231",
abstract = "Textual counterfactual examples explain a prediction by modifying the tokens of an initial instance in order to flip the outcome of a classifier. Even under sparsity constraint, counterfactual generation can lead to numerous changes from the initial text, making the explanation hard to understand. We propose Counterfactual Feature Importance, a method to make non-sparse counterfactual explanations more intelligible. Counterfactual Feature Importance assesses token change importance between an instance to explain and its counterfactual example. We develop two ways of computing Counterfactual Feature Importance, respectively based on classifier gradient computation and counterfactual generator loss evolution during counterfactual search. Then we design a global version of Counterfactual Feature Importance, providing rich information about semantic fields globally impacting classifier predictions. Counterfactual Feature Importance enables to focus on impacting parts of counterfactual explanations, making counterfactual explanations involving numerous changes more understandable.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bhan-etal-2023-enhancing">
<titleInfo>
<title>Enhancing textual counterfactual explanation intelligibility through Counterfactual Feature Importance</title>
</titleInfo>
<name type="personal">
<namePart type="given">Milan</namePart>
<namePart type="family">Bhan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jean-noel</namePart>
<namePart type="family">Vittaut</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nicolas</namePart>
<namePart type="family">Chesneau</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marie-jeanne</namePart>
<namePart type="family">Lesot</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 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anaelia</namePart>
<namePart type="family">Ovalle</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai-Wei</namePart>
<namePart type="family">Chang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ninareh</namePart>
<namePart type="family">Mehrabi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yada</namePart>
<namePart type="family">Pruksachatkun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aram</namePart>
<namePart type="family">Galystan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jwala</namePart>
<namePart type="family">Dhamala</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Apurv</namePart>
<namePart type="family">Verma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Trista</namePart>
<namePart type="family">Cao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anoop</namePart>
<namePart type="family">Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rahul</namePart>
<namePart type="family">Gupta</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>Textual counterfactual examples explain a prediction by modifying the tokens of an initial instance in order to flip the outcome of a classifier. Even under sparsity constraint, counterfactual generation can lead to numerous changes from the initial text, making the explanation hard to understand. We propose Counterfactual Feature Importance, a method to make non-sparse counterfactual explanations more intelligible. Counterfactual Feature Importance assesses token change importance between an instance to explain and its counterfactual example. We develop two ways of computing Counterfactual Feature Importance, respectively based on classifier gradient computation and counterfactual generator loss evolution during counterfactual search. Then we design a global version of Counterfactual Feature Importance, providing rich information about semantic fields globally impacting classifier predictions. Counterfactual Feature Importance enables to focus on impacting parts of counterfactual explanations, making counterfactual explanations involving numerous changes more understandable.</abstract>
<identifier type="citekey">bhan-etal-2023-enhancing</identifier>
<identifier type="doi">10.18653/v1/2023.trustnlp-1.19</identifier>
<location>
<url>https://aclanthology.org/2023.trustnlp-1.19</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>221</start>
<end>231</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Enhancing textual counterfactual explanation intelligibility through Counterfactual Feature Importance
%A Bhan, Milan
%A Vittaut, Jean-noel
%A Chesneau, Nicolas
%A Lesot, Marie-jeanne
%Y Ovalle, Anaelia
%Y Chang, Kai-Wei
%Y Mehrabi, Ninareh
%Y Pruksachatkun, Yada
%Y Galystan, Aram
%Y Dhamala, Jwala
%Y Verma, Apurv
%Y Cao, Trista
%Y Kumar, Anoop
%Y Gupta, Rahul
%S Proceedings of the 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023)
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F bhan-etal-2023-enhancing
%X Textual counterfactual examples explain a prediction by modifying the tokens of an initial instance in order to flip the outcome of a classifier. Even under sparsity constraint, counterfactual generation can lead to numerous changes from the initial text, making the explanation hard to understand. We propose Counterfactual Feature Importance, a method to make non-sparse counterfactual explanations more intelligible. Counterfactual Feature Importance assesses token change importance between an instance to explain and its counterfactual example. We develop two ways of computing Counterfactual Feature Importance, respectively based on classifier gradient computation and counterfactual generator loss evolution during counterfactual search. Then we design a global version of Counterfactual Feature Importance, providing rich information about semantic fields globally impacting classifier predictions. Counterfactual Feature Importance enables to focus on impacting parts of counterfactual explanations, making counterfactual explanations involving numerous changes more understandable.
%R 10.18653/v1/2023.trustnlp-1.19
%U https://aclanthology.org/2023.trustnlp-1.19
%U https://doi.org/10.18653/v1/2023.trustnlp-1.19
%P 221-231
Markdown (Informal)
[Enhancing textual counterfactual explanation intelligibility through Counterfactual Feature Importance](https://aclanthology.org/2023.trustnlp-1.19) (Bhan et al., TrustNLP 2023)
ACL