Two sides of the same coin: Exploiting the impact of identifiers in neural code comprehension
Previous studies have demonstrated that neural code comprehension models are vulnerable to identifier naming. By renaming as few as one identifier in the source code, the models would output completely irrelevant results, indicating that identifiers can be misleading for model prediction. However, i...
Saved in:
Main Authors: | , , , , , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2023
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/9270 https://ink.library.smu.edu.sg/context/sis_research/article/10270/viewcontent/TwoSidesCoin_2023_av.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Singapore Management University |
Language: | English |
id |
sg-smu-ink.sis_research-10270 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-102702024-09-09T07:02:36Z Two sides of the same coin: Exploiting the impact of identifiers in neural code comprehension GAO, Shuzheng GAO, Cuiyun WANG, Chaozheng SUN, Jun LO, David YU, Yue Previous studies have demonstrated that neural code comprehension models are vulnerable to identifier naming. By renaming as few as one identifier in the source code, the models would output completely irrelevant results, indicating that identifiers can be misleading for model prediction. However, identifiers are not completely detrimental to code comprehension, since the semantics of identifier names can be related to the program semantics. Well exploiting the two opposite impacts of identifiers is essential for enhancing the robustness and accuracy of neural code comprehension, and still remains under-explored. In this work, we propose to model the impact of identifiers from a novel causal perspective, and propose a counterfactual reasoning-based framework named CREAM. CREAM explicitly captures the misleading information of identifiers through multi-task learning in the training stage, and reduces the misleading impact by counterfactual inference in the inference stage. We evaluate CREAM on three popular neural code comprehension tasks, including function naming, defect detection and code classification. Experiment results show that CREAM not only significantly outperforms baselines in terms of robustness (e.g., +37.9% on the function naming task at F1 score), but also achieve improved results on the original datasets (e.g., +0.5% on the function naming task at F1 score). 2023-05-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/9270 info:doi/10.1109/ICSE48619.2023.00164 https://ink.library.smu.edu.sg/context/sis_research/article/10270/viewcontent/TwoSidesCoin_2023_av.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Code comprehension Comprehension models Counterfactuals F1 scores Misleading informations Model prediction Multitask learning Neural code Program semantics Source codes Software Engineering |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
Code comprehension Comprehension models Counterfactuals F1 scores Misleading informations Model prediction Multitask learning Neural code Program semantics Source codes Software Engineering |
spellingShingle |
Code comprehension Comprehension models Counterfactuals F1 scores Misleading informations Model prediction Multitask learning Neural code Program semantics Source codes Software Engineering GAO, Shuzheng GAO, Cuiyun WANG, Chaozheng SUN, Jun LO, David YU, Yue Two sides of the same coin: Exploiting the impact of identifiers in neural code comprehension |
description |
Previous studies have demonstrated that neural code comprehension models are vulnerable to identifier naming. By renaming as few as one identifier in the source code, the models would output completely irrelevant results, indicating that identifiers can be misleading for model prediction. However, identifiers are not completely detrimental to code comprehension, since the semantics of identifier names can be related to the program semantics. Well exploiting the two opposite impacts of identifiers is essential for enhancing the robustness and accuracy of neural code comprehension, and still remains under-explored. In this work, we propose to model the impact of identifiers from a novel causal perspective, and propose a counterfactual reasoning-based framework named CREAM. CREAM explicitly captures the misleading information of identifiers through multi-task learning in the training stage, and reduces the misleading impact by counterfactual inference in the inference stage. We evaluate CREAM on three popular neural code comprehension tasks, including function naming, defect detection and code classification. Experiment results show that CREAM not only significantly outperforms baselines in terms of robustness (e.g., +37.9% on the function naming task at F1 score), but also achieve improved results on the original datasets (e.g., +0.5% on the function naming task at F1 score). |
format |
text |
author |
GAO, Shuzheng GAO, Cuiyun WANG, Chaozheng SUN, Jun LO, David YU, Yue |
author_facet |
GAO, Shuzheng GAO, Cuiyun WANG, Chaozheng SUN, Jun LO, David YU, Yue |
author_sort |
GAO, Shuzheng |
title |
Two sides of the same coin: Exploiting the impact of identifiers in neural code comprehension |
title_short |
Two sides of the same coin: Exploiting the impact of identifiers in neural code comprehension |
title_full |
Two sides of the same coin: Exploiting the impact of identifiers in neural code comprehension |
title_fullStr |
Two sides of the same coin: Exploiting the impact of identifiers in neural code comprehension |
title_full_unstemmed |
Two sides of the same coin: Exploiting the impact of identifiers in neural code comprehension |
title_sort |
two sides of the same coin: exploiting the impact of identifiers in neural code comprehension |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2023 |
url |
https://ink.library.smu.edu.sg/sis_research/9270 https://ink.library.smu.edu.sg/context/sis_research/article/10270/viewcontent/TwoSidesCoin_2023_av.pdf |
_version_ |
1814047869962813440 |